Comparison of Two Approaches to Insulin Therapy in Patients With Type 2 Diabetes (IOOX)

• Sponsor: Eli Lilly and Company (Industry)
• Overall Status: Completed
• CT.gov ID: NCT00377858
• Collaborator: (none)
• Enrollment: 484
• Locations: 25
• Arms: 2
• Duration (Months): 25
• Patients Per Site: 19.4
• Patients Per Site Per Month: 0.8

Study Details

Study Description

Brief Summary

A study of patients with type 2 diabetes and inadequate glycemic control on two or more oral antihyperglycemic agents comparing adding insulin lispro mid mixture to the oral antihyperglycemic agents to adding insulin glargine to the oral antihyperglycemic agents.

Condition or Disease	Intervention/Treatment	Phase
Diabetes Mellitus, Type 2	Drug: Insulin lispro mid mixture (MM); Drug: Insulin glargine	Phase 4

Study Design

• Study Type: Interventional
• Actual Enrollment: 484 participants
• Allocation: Randomized
• Intervention Model: Parallel Assignment
• Masking: None (Open Label)
• Primary Purpose: Treatment
• Official Title: Comparison of Two Approaches to Basal-Bolus Insulin Therapy in Patients With Type 2 Diabetes and Inadequate Glycemic Control on Oral Therapy: Comparison of Premixed Insulin Lispro Mid Mixture With Separate Basal and Bolus Insulin Injections
• Study Start Date: Aug 1, 2006
• Actual Primary Completion Date: Sep 1, 2008
• Actual Study Completion Date: Sep 1, 2008

Arms and Interventions

Arm	Intervention/Treatment
Experimental: Insulin Lispro Mid Mixture Insulin lispro mid mixture (MM) up to three times a day (TID)	Drug: Insulin lispro mid mixture (MM); Patient specific adjusted dose, three times a day (TID), subcutaneous (SC) injection x 36 weeks; Other Names: Humalog Mid Mix
Active Comparator: Insulin Glargine Insulin glargine daily with insulin lispro at mealtime (up to 3 injections) as needed.	Drug: Insulin glargine; Patient specific adjusted dose, every day (QD), subcutaneous (SC) injection x 36 weeks

Outcome Measures

Primary Outcome Measures

1. Hemoglobin A1c (HbA1c) at 36 Week Endpoint [36 weeks]

   Level of hemoglobin A1c at endpoint.

Secondary Outcome Measures

1. Hemoglobin A1c (HbA1c) at Interval Visits [12, 24, and 36 weeks]

   Levels of HbA1c at 12 weeks and 24 weeks and 36 weeks.

2. Percentage of Patients Who Achieved Hemoglobin A1c Less Than or Equal to 6.5%, Greater Than 6.5%, Less Than 7%, Greater Than or Equal to 7%, Less Than or Equal to 7%, and Greater Than 7% at Interval Visits and Endpoint [12-24-36 weeks]

3. 7-point Self-monitored Blood Glucose Profiles [Baseline, 12-24-36 weeks]

   Actual daily mean blood glucose levels at specified time points.

4. Glycemic Variability [Baseline, 12-24-36 weeks]

   Glycemic variability was measured by mean blood glucose value (M-value), which was the mean of the intra-days self-monitoring blood glucose values, and by the mean of daily difference (MODD), which was the mean of the between-days self-monitored blood glucose values.

5. Number of Patients With at Least One Self-reported Hypoglycemic Episode, Including Nocturnal (and Non-nocturnal) Hypoglycemia [Baseline to 36 Weeks]

   Hypoglycemic episode defined: any time patient felt that he/she was experiencing a sign or symptom associated with hypoglycemia, or had old Roche blood glucose level <70 mg/dL even if not associated with signs, symptoms, or treatment consistent with current guidelines. Nocturnal hypoglycemia defined: any hypoglycemic event that occurred between bedtime and waking. Non-nocturnal hypoglycemia defined: any hypoglycemic event that occurred between waking and bedtime. Overall episodes: those that occurred at any time during the post-randomization visits. Endpoint: last visit interval based on LOCF.

6. 30-Day Adjusted Rates of Self-Reported Hypoglycemic Episodes (Including Nocturnal and Non-Nocturnal) [Baseline to 36 Weeks]

   Hypoglycemic episode defined: any time patient felt that he/she was experiencing a sign or symptom associated with hypoglycemia, or had old Roche blood glucose level <70 mg/dL even if not associated with signs, symptoms, or treatment consistent with current guidelines. Nocturnal hypoglycemia defined: any hypoglycemic event that occurred between bedtime and waking. Non-nocturnal hypoglycemia defined: any hypoglycemic event that occurred between waking and bedtime. Overall episodes: those that occurred at any time during the post-randomization visits. Endpoint: last visit interval based on LOCF.

7. Number of Patients With at Least One Severe Hypoglycemia Episode [Baseline to 36 Weeks]

   Severe hypoglycemia was defined as hypoglycemic event that meets at least one of the following criteria: not capable of treating self and blood glucose <2.8 millimoles per liter (mmol/L); not capable of treating self, blood glucose is missing and prompt recovery after oral carbohydrate or glucagon or intravenous glucose; hypoglycemic event outcome was coma, hopitalization, emergency room visit, or automobile accident. The overall category is a severe hypoglycemic event that occurred at any time during the post-randomization visits. Endpoint: last visit interval based on LOCF.

8. Endpoint Insulin Dose Per Body Weight; Total, Basal, and Prandial [36 Weeks]

   Total daily insulin dose adjusted for body weight (Units of insulin per kilogram per day [U/kg/day]) was assessed. Basal insulin is the amount of insulin required to manage normal daily blood glucose fluctuations. Prandial insulin is taken at meal time. Insulin glargine is a basal insulin and insulin lispro is a prandial insulin. Insulin lispro mid-mix is a 50/50 mixture of a basal insulin and insulin lispro. Endpoint: last visit interval based on LOCF.

9. Endpoint Insulin Dose; Total, Basal, and Prandial [36 Weeks]

   Total daily insulin dose (Units of insulin per day [U/day]) was assessed. Basal insulin is the amount of insulin required to manage normal daily blood glucose fluctuations. Prandial insulin is taken at meal time. Insulin glargine is a basal insulin and insulin lispro is a prandial insulin. Insulin lispro mid-mix is a 50/50 mixture of a basal insulin and insulin lispro. Endpoint: last visit interval based on LOCF.

10. Number of Insulin Injections Per Day [Weeks 12, 24, 30, 36]

11. Change From Baseline in Absolute Body Weight at 36 Week Endpoint [Baseline, 36 Weeks]

    Change in body weight was calculated as weight at endpoint (last observation carried forward) minus weight at baseline.

Eligibility Criteria

Criteria

• Ages Eligible for Study: 30 Years to 80 Years
• Sexes Eligible for Study: All
• Accepts Healthy Volunteers: No

Inclusion Criteria:

• Have type 2 diabetes

• Have been receiving oral antihyperglycemic medications (OAM) without insulin including at least two of the following at maximally tolerated doses, AND meet the minimum dosing criteria shown: Metformin 1500 mg/day, Sulfonylurea 1/2 the maximum daily dose, according to package insert, Thiazolidinedione (TZD) 30 mg/day pioglitazone or 4 mg/day rosiglitazone. The OAMs also must be used in accordance with the product label

• Have a hemoglobin A1c greater than or equal to 7.5% and less than or equal to 12.0%.

Exclusion Criteria:

• Are taking a TZD dose greater than what is indicated in combination with insulin according to the TZD label.

• Are taking any other glucose-lowering agents not mentioned in Inclusion Criterion.

• Have taken acarbose, miglitol, pramlintide, exenatide, repaglinide, or nateglinide in the past 6 weeks or for a total of 30 days or more in the last 24 weeks.

• Have a body mass index greater than 40 kg/m2.

• Have had more than one episode of severe hypoglycemia in the last24 weeks

• Are pregnant, intend to be pregnant during the course of the study or are breastfeeding

• Have clinically significant cardiac, renal, hematologic, oncologic, or hepatic disease

Contacts and Locations

Locations

	Site	City	State	Country	Postal Code
1	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Keswick	South Australia	Australia	5035
2	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Fitzroy	Victoria	Australia	3065
3	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Fremantle	Western Australia	Australia	6160
4	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	London	Ontario	Canada	N6A 4L2
5	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Granby	Quebec	Canada	J2G 1T7
6	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Sherbrooke	Quebec	Canada	J1G 5K2
7	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Regina	Saskatchewan	Canada	S4P 0W5
8	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Mantes La Jolie		France	78200
9	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Menton		France	06500
10	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Pau		France	64000
11	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Poitiers		France	86000
12	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Toulouse		France	31082
13	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Goyang-Si/Kyunggi-Do		Korea, Republic of	410-719
14	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Kwang Ju		Korea, Republic of	501-757
15	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Seoul		Korea, Republic of	110-746
16	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Chihuahua		Mexico	31238
17	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Guadalajara		Mexico	44620
18	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Arkhangelsk		Russian Federation	163045
19	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Moscow		Russian Federation	117036
20	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Saint Petersburg		Russian Federation	193257
21	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Alicante		Spain	03114
22	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Almeria		Spain	04001
23	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Barcelona		Spain	08017
24	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Bilbao		Spain	48013
25	For additional information regarding investigative sites for this trial, contact 1-877-CTLILLY (1-877-285-4559, 1-317-615-4559) Mon - Fri from 9 AM to 5 PM Eastern Time (UTC/GMT - 5 hours, EST), or speak with your personal physician.	Malaga		Spain	29010

Sponsors and Collaborators

• Eli Lilly and Company

Investigators

• Study Director: Call 1-877-CTLILLY (1-877-285-4559 or 1-317-615-4559 Mon-Fri 9 AM - 5 PM Eastern time (UTC/GMT - 5 hours, EST), Eli Lilly and Company

Study Documents (Full-Text)

None provided.

More Information

Additional Information:

• Lilly Clinical Trial Registry

Publications

None provided.

• ClinicalTrials.gov Identifier: NCT00377858
• Other Study ID Numbers: 10936; F3Z-MC-IOOX
• First Posted: Sep 19, 2006
• Last Update Posted: Dec 9, 2009
• Last Verified: Dec 1, 2009

• Keywords provided by , ,: diabetes; type 2
• Additional relevant MeSH terms: Diabetes Mellitus; Diabetes Mellitus, Type 2; Insulin; Insulin, Globin Zinc; Insulin Glargine; Insulin Lispro

Study Results

Participant Flow

Pre-assignment Detail	636 patients were screened; 128 were screen failures; 24 discontinued prior to randomization; 484 patients were randomized. Baseline Characteristics are provided for the Intent to Treat Population, which was defined as patients with baseline and at least one post-baseline value.

Arm/Group Title	Insulin Lispro Mid Mixture	Insulin Glargine
Arm/Group Description	Insulin lispro mid mixture (MM) up to three times a day (TID)	Insulin glargine daily with insulin lispro at mealtime (up to 3 injections) as needed.
Period Title: Overall Study
STARTED	242	242
Had at Least One Post-Baseline Value	239	240
COMPLETED	211	215
NOT COMPLETED	31	27

Baseline Characteristics

Arm/Group Title	Insulin Lispro Mid Mixture	Insulin Glargine	Total
Arm/Group Description	Insulin lispro mid mixture (MM) up to three times a day (TID)	Insulin glargine daily with insulin lispro at mealtime (up to 3 injections) as needed.	Total of all reporting groups
Overall Participants	239	240	479
Age (years) [Mean (Standard Deviation) ]
Mean (Standard Deviation) [years]	58.40 (9.21)	59.45 (9.87)	58.92 (9.55)
Sex: Female, Male (Count of Participants)
Female	131 54.8%	120 50%	251 52.4%
Male	108 45.2%	120 50%	228 47.6%
Region of Enrollment (participants) [Number]
Australia	18 7.5%	19 7.9%	37 7.7%
Canada	27 11.3%	24 10%	51 10.6%
France	21 8.8%	19 7.9%	40 8.4%
Greece	19 7.9%	20 8.3%	39 8.1%
India	26 10.9%	29 12.1%	55 11.5%
Korea, Republic of	24 10%	25 10.4%	49 10.2%
Mexico	51 21.3%	47 19.6%	98 20.5%
Russian Federation	24 10%	26 10.8%	50 10.4%
Spain	29 12.1%	31 12.9%	60 12.5%
Body Mass Index (BMI) Group (participants) [Number]
BMI <25 kg/m^2	34 14.2%	48 20%	82 17.1%
BMI ≥25 kg/m^2 and <30 kg/m^2	106 44.4%	103 42.9%	209 43.6%
BMI ≥30 kg/m^2	99 41.4%	89 37.1%	188 39.2%
Duration of Diabetes Group (participants) [Number]
<5 years	28 11.7%	27 11.3%	55 11.5%
≥5 years and ≤10 years	95 39.7%	100 41.7%	195 40.7%
>10 years	116 48.5%	113 47.1%	229 47.8%
Hemoglobin A1c (HbA1c) Group (participants) [Number]
≤8.5% HbA1c	60 25.1%	67 27.9%	127 26.5%
>8.5% HbA1c	179 74.9%	173 72.1%	352 73.5%
Race/Ethnicity (participants) [Number]
African	2 0.8%	1 0.4%	3 0.6%
Caucasian	134 56.1%	136 56.7%	270 56.4%
East Asian	25 10.5%	27 11.3%	52 10.9%
Hispanic	51 21.3%	46 19.2%	97 20.3%
West Asian (Indian sub-continent)	27 11.3%	30 12.5%	57 11.9%
Body Height (centimeters (cm)) [Mean (Standard Deviation) ]
Mean (Standard Deviation) [centimeters (cm)]	163.27 (9.27)	164.68 (9.48)	163.98 (9.39)
Body Mass Index (BMI) (kilograms per square meter (kg/m^2)) [Mean (Standard Deviation) ]
Mean (Standard Deviation) [kilograms per square meter (kg/m^2)]	29.51 (4.76)	29.00 (4.55)	29.25 (4.66)
Body Weight (kilograms (kg)) [Mean (Standard Deviation) ]
Mean (Standard Deviation) [kilograms (kg)]	78.85 (15.50)	78.87 (15.10)	78.86 (15.28)
Duration of Diabetes (years) [Mean (Standard Deviation) ]
Mean (Standard Deviation) [years]	10.86 (5.54)	11.79 (7.29)	11.33 (6.48)
Hemoglobin A1c (HbA1c) (percent HbA1c) [Mean (Standard Deviation) ]
Mean (Standard Deviation) [percent HbA1c]	9.55 (1.18)	9.39 (1.17)	9.47 (1.18)

Outcome Measures

1. Primary Outcome

Title	Hemoglobin A1c (HbA1c) at 36 Week Endpoint
Description	Level of hemoglobin A1c at endpoint.
Time Frame	36 weeks

Outcome Measure Data

Analysis Population Description

Number of participants in the per-protocol population. Last observation carried forward.

Arm/Group Title	Insulin Lispro Mid Mixture	Insulin Glargine
Arm/Group Description	Insulin lispro mid mixture (MM) up to three times a day (TID)	Insulin glargine daily with insulin lispro at mealtime (up to 3 injections) as needed.
Measure Participants	188	195
Least Squares Mean (Standard Error) [percent HbA1c]	7.66 (0.37)	7.49 (0.36)

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments	Assuming 15% drop-out rate after randomization, remaining 213 patients in each treatment group would allow confirmation of noninferiority with no treatment difference and a noninferiority limit of 0.3% using upper limit of 2-sided confidence interval at significance level of 0.05 with 80% power.
	Type of Statistical Test	Non-Inferiority or Equivalence
	Comments	Noninferiority margin of 0.3% based on prior studies indicating an HbA1c difference of 0.6% in patients treated with lispro and sulfonylurea compared with those treated with sulfonylurea and metformin.
Statistical Test of Hypothesis	p-Value	0.097
	Comments
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline HbA1c Stratum + Sulfonylurea stratum + Country + Baseline HbA1c Stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.17
	Confidence Interval	() 95%; -0.03 to 0.37
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for (Insulin Lispro Mid Mixture minus Insulin Glargine).

2. Secondary Outcome

Title	Hemoglobin A1c (HbA1c) at Interval Visits
Description	Levels of HbA1c at 12 weeks and 24 weeks and 36 weeks.
Time Frame	12, 24, and 36 weeks

Outcome Measure Data

Analysis Population Description

Number of randomized patients with baseline and at least one post-baseline value. Intent to treat population.

Arm/Group Title	Insulin Lispro Mid Mixture	Insulin Glargine
Arm/Group Description	Insulin lispro mid mixture (MM) up to three times a day (TID)	Insulin glargine daily with insulin lispro at mealtime (up to 3 injections) as needed.
Measure Participants	222	222
12 Week Interval (n=222, n=222)	8.02 (0.32)	7.85 (0.31)
24 Week Interval (n=206, n=217)	7.53 (0.33)	7.36 (0.32)
36 Week Interval (n=202, n=208)	7.53 (0.37)	7.44 (0.37)

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.047
	Comments	P-value for 12 Week Interval.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline HbA1c + Baseline HbA1c stratum + Sulfonylurea stratum + Country.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.16
	Confidence Interval	() 95%; 0.00 to 0.33
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.043
	Comments	P-value for 24 Week Interval.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline HbA1c + Baseline HbA1c stratum + Sulfonylurea stratum + Country.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.18
	Confidence Interval	() 95%; 0.01 to 0.35
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.343
	Comments	P-value for 36 Week Interval.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline HbA1c + Baseline HbA1c stratum + Sulfonylurea stratum + Country.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.10
	Confidence Interval	() 95%; -0.10 to 0.29
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

3. Secondary Outcome

Title	Percentage of Patients Who Achieved Hemoglobin A1c Less Than or Equal to 6.5%, Greater Than 6.5%, Less Than 7%, Greater Than or Equal to 7%, Less Than or Equal to 7%, and Greater Than 7% at Interval Visits and Endpoint
Time Frame	12-24-36 weeks

Outcome Measure Data

Analysis Population Description

Number of randomized patients with baseline and at least one post-baseline value. Intent to treat population. If the patient completed the trial, the endpoint was 36 weeks (Week 36), but if the patient dropped out of the study early, the last observation carried forward (LOCF) was considered as the endpoint (Endpoint).

Arm/Group Title	Insulin Lispro Mid Mixture	Insulin Glargine
Arm/Group Description	Insulin lispro mid mixture (MM) up to three times a day (TID)	Insulin glargine daily with insulin lispro at mealtime (up to 3 injections) as needed.
Measure Participants	234	235
Week 12: HbA1c ≤7.0% (n=222, n=222)	21.2 8.9%	25.2 10.5%
Week 12: HbA1c >7.0% (n=222, n=222)	78.8 33%	74.8 31.2%
Week 12: HbA1c <7.0% (n=222, n=222)	18.9 7.9%	21.6 9%
Week 12: HbA1c ≥7.0% (n=222, n=222)	81.1 33.9%	78.4 32.7%
Week 12: HbA1c ≤6.5% (n=222, n=222)	8.1 3.4%	10.8 4.5%
Week 12: HbA1c >6.5% (n=222, n=222)	91.9 38.5%	89.2 37.2%
Week 24: HbA1c ≤7.0% (n=206, n=217)	39.8 16.7%	47.0 19.6%
Week 24: HbA1c >7.0% (n=206, n=217)	60.2 25.2%	53.0 22.1%
Week 24: HbA1c <7.0% (n=206, n=217)	35.4 14.8%	38.7 16.1%
Week 24: HbA1c ≥7.0% (n=206, n=217)	64.6 27%	61.3 25.5%
Week 24: HbA1c ≤6.5% (n=206, n=217)	16.0 6.7%	21.7 9%
Week 24: HbA1c >6.5% (n=206, n=217)	84.0 35.1%	78.3 32.6%
Week 36: HbA1c ≤7.0% (n=202, n=208)	39.6 16.6%	44.7 18.6%
Week 36: HbA1c >7.0% (n=202, n=208)	60.4 25.3%	55.3 23%
Week 36: HbA1c <7.0% (n=202, n=208)	37.6 15.7%	40.4 16.8%
Week 36: HbA1c ≥7.0% (n=202, n=208)	62.4 26.1%	59.6 24.8%
Week 36: HbA1c ≤6.5% (n=202, n=208)	15.3 6.4%	19.2 8%
Week 36: HbA1c >6.5% (n=202, n=208)	84.7 35.4%	80.8 33.7%
Endpoint (LOCF): HbA1c ≤7.0% (n=234, n=235)	36.8 15.4%	43.0 17.9%
Endpoint (LOCF): HbA1c >7.0% (n=234, n=235)	63.2 26.4%	57.0 23.8%
Endpoint (LOCF): HbA1c <7.0% (n=234, n=235)	35.0 14.6%	39.1 16.3%
Endpoint (LOCF): HbA1c ≥7.0% (n=234, n=235)	65.0 27.2%	60.9 25.4%
Endpoint (LOCF): HbA1c ≤6.5% (n=234, n=235)	13.2 5.5%	19.1 8%
Endpoint (LOCF): HbA1c >6.5% (n=234, n=235)	86.8 36.3%	80.9 33.7%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.432
	Comments	P-value for Week 12: HbA1c ≤7.0%.
	Method	Regression, Logistic
	Comments	Logistic Regression Model: Variable=Treatment + Baseline HbA1c + Country + Sulfonylurea Stratum.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.602
	Comments	P-value for 12 Week: HbA1c <7.0%.
	Method	Regression, Logistic
	Comments	Logistic Regression Model: Variable=Treatment + Baseline HbA1c + Country + Sulfonylurea Stratum.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.370
	Comments	P-value for 12 Week: HbA1c ≤6.5%.
	Method	Regression, Logistic
	Comments	Logistic Regression Model: Variable=Treatment + Baseline HbA1c + Country + Sulfonylurea Stratum.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.198
	Comments	P-value for 24 Week: HbA1c ≤7.0%.
	Method	Regression, Logistic
	Comments	Logistic Regression Model: Variable=Treatment + Baseline HbA1c + Country + Sulfonylurea Stratum.

Statistical Analysis 5

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.636
	Comments	P-value for 24 Week: HbA1c <7.0%.
	Method	Regression, Logistic
	Comments	Logistic Regression Model: Variable=Treatment + Baseline HbA1c + Country + Sulfonylurea Stratum.

Statistical Analysis 6

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.185
	Comments	P-value for 24 Week: HbA1c ≤6.5%.
	Method	Regression, Logistic
	Comments	Logistic Regression Model: Variable=Treatment + Baseline HbA1c + Country + Sulfonylurea Stratum.

Statistical Analysis 7

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.393
	Comments	P-value for 36 Week: HbA1c ≤7.0%.
	Method	Regression, Logistic
	Comments	Logistic Regression Model: Variable=Treatment + Baseline HbA1c + Country + Sulfonylurea Stratum.

Statistical Analysis 8

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.739
	Comments	P-value for 36 Week: HbA1c <7.0%.
	Method	Regression, Logistic
	Comments	Logistic Regression Model: Variable=Treatment + Baseline HbA1c + Country + Sulfonylurea Stratum.

Statistical Analysis 9

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.396
	Comments	P-value for 36 Week: HbA1c ≤6.5%.
	Method	Regression, Logistic
	Comments	Logistic Regression Model: Variable=Treatment + Baseline HbA1c + Country + Sulfonylurea Stratum.

Statistical Analysis 10

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.227
	Comments	P-value for Endpoint (LOCF): HbA1c ≤7.0%.
	Method	Regression, Logistic
	Comments	Logistic Regression Model: Variable=Treatment + Baseline HbA1c + Country + Sulfonylurea Stratum.

Statistical Analysis 11

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.482
	Comments	P-value for Endpoint (LOCF): HbA1c <7.0%.
	Method	Regression, Logistic
	Comments	Logistic Regression Model: Variable=Treatment + Baseline HbA1c + Country + Sulfonylurea Stratum.

Statistical Analysis 12

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.108
	Comments	P-value for Endpoint (LOCF): HbA1c ≤6.5%.
	Method	Regression, Logistic
	Comments	Logistic Regression Model: Variable=Treatment + Baseline HbA1c + Country + Sulfonylurea Stratum.

4. Secondary Outcome

Title	7-point Self-monitored Blood Glucose Profiles
Description	Actual daily mean blood glucose levels at specified time points.
Time Frame	Baseline, 12-24-36 weeks

Outcome Measure Data

Analysis Population Description

Number of randomized patients with baseline and at least one post-baseline value. Intent to treat population. If the patient completed the trial, the endpoint was 36 weeks (Week 36), but if the patient dropped out of the study early, the last observation carried forward (LOCF) was considered as the endpoint (Endpoint).

Arm/Group Title	Insulin Lispro Mid Mixture	Insulin Glargine
Arm/Group Description	Insulin lispro mid mixture (MM) up to three times a day (TID)	Insulin glargine daily with insulin lispro at mealtime (up to 3 injections) as needed.
Measure Participants	239	240
Baseline: Morning Pre-Meal (n=228, n=231)	10.22 (0.91)	9.64 (0.90)
Baseline: Morning Postprandial Meal (n=213, n=222)	12.49 (1.18)	12.39 (1.16)
Baseline: Midday Pre-Meal (n=227, n=230)	10.15 (1.07)	9.62 (1.05)
Baseline: Midday Postprandial Meal (n=215, n=223)	11.40 (1.12)	11.22 (1.10)
Baseline: Evening Pre-Meal (n=227, n=229)	10.38 (1.09)	10.31 (1.07)
Baseline: Evening Postprandial Meal (n=226,n= 229)	11.87 (1.05)	11.76 (1.03)
Baseline: 0300 Hours (n=205, n=202)	9.37 (1.08)	9.17 (1.06)
12 Week: Morning Pre-Meal (n=218, n=226)	7.26 (0.64)	6.68 (0.63)
12 Week: Morning Postprandial Meal (n=192, n=206)	11.56 (0.94)	11.28 (0.92)
12 Week: Midday Pre-Meal (n=215, n=220)	8.63 (0.76)	8.10 (0.75)
12 Week: Midday Postprandial Meal (n=196, n=209)	9.92 (0.91)	9.43 (0.89)
12 Week: Evening Pre-Meal (n=216, n=221)	8.75 (0.80)	7.98 (0.79)
12 Week: Evening Postprandial Meal (n=213, n= 223)	8.23 (0.90)	9.40 (0.89)
12 Week: 0300 Hours (n=185, n=187)	6.97 (0.79)	7.12 (0.78)
24 Week: Morning Pre-Meal (n=210, n=213)	6.72 (0.64)	6.07 (0.63)
24 Week: Morning Postprandial Meal (n=193, n=201)	9.87 (0.91)	9.45 (0.89)
24 Week: Midday Pre-Meal (n=209, n=213)	7.99 (0.79)	7.82 (0.77)
24 Week: Midday Postprandial Meal (n=196, n=203)	9.38 (0.90)	9.15 (0.88)
24 Week: Evening Pre-Meal (n=209, n=211)	7.50 (0.81)	6.94 (0.79)
24 Week: Evening Postprandial Meal (n=207, n= 207)	8.21 (0.82)	8.69 (0.81)
24 Week: 0300 Hours (n=181, n=184)	7.14 (0.71)	7.18 (0.70)
36 Week: Morning Pre-Meal (n=200, n=210)	6.96 (0.67)	6.48 (0.65)
36 Week: Morning Postprandial Meal (n=179, n=194)	10.09 (0.92)	10.29 (0.90)
36 Week: Midday Pre-Meal (n=199, n=210)	7.33 (0.77)	7.46 (0.76)
36 Week: Midday Postprandial Meal (n=182, n=195)	10.31 (0.92)	10.30 (0.91)
36 Week: Evening Pre-Meal (n=199, n=207)	8.13 (0.81)	7.97 (0.79)
36 Week: Evening Postprandial Meal (n=195, n= 225)	9.46 (0.85)	9.85 (0.83)
36 Week: 0300 Hours (n=167, n=177)	8.18 (1.06)	8.25 (1.05)
Endpoint: Morning Pre-Meal (n=228, n=231)	6.98 (0.68)	6.53 (0.67)
Endpoint: Morning Postprandial Meal (n=213, n=222)	9.98 (0.94)	10.29 (0.92)
Endpoint: Midday Pre-Meal (n=227, n=230)	7.36 (0.80)	7.52 (0.79)
Endpoint: Midday Postprandial Meal (n=215, n=223)	10.28 (0.91)	10.34 (0.90)
Endpoint: Evening Pre-Meal (n=227, n=229)	8.15 (0.81)	7.98 (0.80)
Endpoint: Evening Postprandial Meal (n=226,n= 229)	9.27 (0.85)	9.84 (0.83)
Endpoint: 0300 Hours (n=205, n=202)	7.78 (0.75)	7.89 (0.74)

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.014
	Comments	P-value for Baseline: Morning Pre-Meal.
	Method	ANOVA
	Comments	ANOVA Model: Variable=Treatment + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.58
	Confidence Interval	() 95%; 0.12 to 1.04
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.735
	Comments	P-value for Baseline: Morning Postprandial Meal.
	Method	ANOVA
	Comments	ANOVA Model: Variable=Treatment + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.11
	Confidence Interval	() 95%; -0.51 to 0.72
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.051
	Comments	P-value for Baseline: Midday Pre-Meal.
	Method	ANOVA
	Comments	ANOVA Model: Variable=Treatment + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.54
	Confidence Interval	() 95%; -0.00 to 1.08
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.536
	Comments	P-value for Baseline: Midday Postprandial Meal.
	Method	ANOVA
	Comments	ANOVA Model: Variable=Treatment + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.18
	Confidence Interval	() 95%; -0.39 to 0.75
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 5

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.793
	Comments	P-value for Baseline: Evening Pre-Meal.
	Method	ANOVA
	Comments	ANOVA Model: Variable=Treatment + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.07
	Confidence Interval	() 95%; -0.48 to 0.62
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 6

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.693
	Comments	P-value for Baseline: Evening Postprandial Meal.
	Method	ANOVA
	Comments	ANOVA Model: Variable=Treatment + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.11
	Confidence Interval	() 95%; -0.43 to 0.64
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 7

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.499
	Comments	P-value for Baseline: 0300 Hours.
	Method	ANOVA
	Comments	ANOVA Model: Variable=Treatment + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.20
	Confidence Interval	() 95%; -0.38 to 0.77
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 8

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	P-value for 12 Week: Morning Pre-Meal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.59
	Confidence Interval	() 95%; 0.26 to 0.92
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 9

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.278
	Comments	P-value for 12 Week: Morning Postprandial Meal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.28
	Confidence Interval	() 95%; -0.22 to 0.78
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 10

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.009
	Comments	P-value for 12 Week: Midday Pre-Meal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.53
	Confidence Interval	() 95%; 0.14 to 0.93
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 11

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.049
	Comments	P-value for 12 Week: Midday Postprandial Meal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.49
	Confidence Interval	() 95%; 0.00 to 0.97
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 12

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	P-value for 12 Week: Evening Pre-Meal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.77
	Confidence Interval	() 95%; 0.35 to 1.18
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 13

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	P-value for 12 Week: Evening Postprandial Meal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-1.17
	Confidence Interval	() 95%; -1.63 to -0.70
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 14

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.506
	Comments	P-value for 12 Week: 0300 Hours.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-0.15
	Confidence Interval	() 95%; -0.59 to 0.29
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 15

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	P-value for 24 Week: Morning Pre-Meal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.65
	Confidence Interval	() 95%; 0.31 to 0.99
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 16

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.094
	Comments	P-value for 24 Week: Morning Postprandial Meal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.42
	Confidence Interval	() 95%; -0.07 to 0.91
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 17

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.408
	Comments	P-value for 24 Week: Midday Pre-Meal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.17
	Confidence Interval	() 95%; -0.24 to 0.59
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 18

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.351
	Comments	P-value for 24 Week: Midday Postprandial Meal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.23
	Confidence Interval	() 95%; -0.25 to 0.71
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 19

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.010
	Comments	P-value for 24 Week: Evening Pre-Meal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.56
	Confidence Interval	() 95%; 0.13 to 0.98
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 20

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.029
	Comments	P-value for 24 Week: Evening Postprandial Meal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-0.48
	Confidence Interval	() 95%; -0.92 to -0.05
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 21

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.816
	Comments	P-value for 24 Week: 0300 Hours.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-0.05
	Confidence Interval	() 95%; -0.45 to 0.35
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 22

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.008
	Comments	P-value for 36 Week: Morning Pre-Meal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.49
	Confidence Interval	() 95%; 0.13 to 0.85
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 23

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.438
	Comments	P-value for 36 Week: Morning Postprandial Meal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-0.20
	Confidence Interval	() 95%; -0.71 to 0.31
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 24

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.548
	Comments	P-value for 36 Week: Midday Pre-Meal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-0.13
	Confidence Interval	() 95%; -0.54 to 0.29
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 25

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.964
	Comments	P-value for 36 Week: Midday Postprandial Meal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.01
	Confidence Interval	() 95%; -0.50 to 0.52
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 26

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.453
	Comments	P-value for 36 Week: Evening Pre-Meal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.16
	Confidence Interval	() 95%; -0.27 to 0.60
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 27

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.096
	Comments	P-value for 36 Week: Evening Postprandial Meal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-0.39
	Confidence Interval	() 95%; -0.84 to 0.07
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 28

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.770
	Comments	P-value for 36 Week: 0300 Hours.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-0.07
	Confidence Interval	() 95%; -0.51 to 0.37
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 29

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.010
	Comments	P-value for Endpoint: Morning Pre-Meal. (last observation carried forward)
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.45
	Confidence Interval	() 95%; 0.11 to 0.80
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 30

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.201
	Comments	P-value for Endpoint: Morning Postprandial Meal. (last observation carried forward)
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-0.31
	Confidence Interval	() 95%; -0.80 to 0.17
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 31

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.430
	Comments	P-value for Endpoint: Midday Pre-Meal. (last observation carried forward)
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-0.16
	Confidence Interval	() 95%; -0.57 to 0.24
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 32

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.814
	Comments	P-value for Endpoint: Midday Postprandial Meal. (last observation carried forward)
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-0.06
	Confidence Interval	() 95%; -0.52 to 0.41
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 33

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.427
	Comments	P-value for Endpoint: Evening Pre-Meal. (last observation carried forward)
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.17
	Confidence Interval	() 95%; -0.24 to 0.57
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 34

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.009
	Comments	P-value for Endpoint: Evening Postprandial Meal. (last observation carried forward)
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-0.57
	Confidence Interval	() 95%; -1.00 to -0.14
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 35

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.569
	Comments	P-value for Endpoint: 0300 Hours. (last observation carried forward)
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-0.12
	Confidence Interval	() 95%; -0.51 to 0.28
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

5. Secondary Outcome

Title	Glycemic Variability
Description	Glycemic variability was measured by mean blood glucose value (M-value), which was the mean of the intra-days self-monitoring blood glucose values, and by the mean of daily difference (MODD), which was the mean of the between-days self-monitored blood glucose values.
Time Frame	Baseline, 12-24-36 weeks

Outcome Measure Data

Analysis Population Description

Number of randomized patients with baseline and at least one post-baseline value. Intent to treat population. If the patient completed the trial, the endpoint was 36 weeks (Week 36), but if the patient dropped out of the study early, the last observation carried forward (LOCF) was considered as the endpoint (Endpoint).

Arm/Group Title	Insulin Lispro Mid Mixture	Insulin Glargine
Arm/Group Description	Insulin lispro mid mixture (MM) up to three times a day (TID)	Insulin glargine daily with insulin lispro at mealtime (up to 3 injections) as needed.
Measure Participants	239	240
Baseline MODD (n=226, n=229)	2.09 (0.33)	2.07 (0.32)
Baseline M-Value (n=228, n=231)	48.02 (14.73)	46.75 (14.47)
12 Week MODD (n=214, n=221)	1.51 (0.31)	1.54 (0.31)
12 Week M-Value (n=218, n=226)	30.07 (7.24)	28.49 (7.12)
24 Week MODD (n=208, n=211)	1.57 (0.33)	1.66 (0.32)
24 Week M-Value (n=210, n=213)	22.67 (7.05)	20.14 (6.93)
36 Week MODD (n=198, n=206)	1.55 (0.32)	1.60 (0.32)
36 Week M-Value (n=200, n=210)	26.46 (7.26)	27.41 (7.12)
Endpoint MODD (n=226, n=229)	1.51 (0.32)	1.61 (0.31)
Endpoint M-Value (n=228, n=231)	26.27 (7.41)	27.99 (7.28)

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.776
	Comments	P-value for Baseline MODD.
	Method	ANOVA
	Comments	ANOVA Model: Variable=Treatment + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.02
	Confidence Interval	() 95%; -0.14 to 0.19
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.737
	Comments	P-value for Baseline M-Value.
	Method	ANOVA
	Comments	ANOVA Model: Variable=Treatment + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	1.27
	Confidence Interval	() 95%; -6.14 to 8.68
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.701
	Comments	P-value for 12 Week MODD.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-0.03
	Confidence Interval	() 95%; -0.19 to 0.13
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.400
	Comments	P-value for 12 Week M-Value.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	1.58
	Confidence Interval	() 95%; -2.11 to 5.27
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 5

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.321
	Comments	P-value for 24 Week MODD.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-0.09
	Confidence Interval	() 95%; -0.26 to 0.09
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 6

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.179
	Comments	P-value for 24 Week M-Value.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	2.52
	Confidence Interval	() 95%; -1.16 to 6.21
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 7

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.573
	Comments	P-value for 36 Week MODD.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-0.05
	Confidence Interval	() 95%; -0.22 to 0.12
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 8

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.629
	Comments	P-value for 36 Week M-Value.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-0.95
	Confidence Interval	() 95%; -4.79 to 2.90
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 9

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.250
	Comments	P-value for Endpoint MODD. (last observation carried forward)
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-0.09
	Confidence Interval	() 95%; -0.26 to 0.07
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 10

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.362
	Comments	P-value for Endpoint M-Value.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-1.72
	Confidence Interval	() 95%; -5.44 to 1.99
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

6. Secondary Outcome

Title	Number of Patients With at Least One Self-reported Hypoglycemic Episode, Including Nocturnal (and Non-nocturnal) Hypoglycemia
Description	Hypoglycemic episode defined: any time patient felt that he/she was experiencing a sign or symptom associated with hypoglycemia, or had old Roche blood glucose level <70 mg/dL even if not associated with signs, symptoms, or treatment consistent with current guidelines. Nocturnal hypoglycemia defined: any hypoglycemic event that occurred between bedtime and waking. Non-nocturnal hypoglycemia defined: any hypoglycemic event that occurred between waking and bedtime. Overall episodes: those that occurred at any time during the post-randomization visits. Endpoint: last visit interval based on LOCF.
Time Frame	Baseline to 36 Weeks

Outcome Measure Data

Analysis Population Description

Number of randomized patients with baseline and at least one post-baseline value. Intent to treat population. The last visit of trial interval was used to calculate hypoglycemic episodes at Endpoint if patient completes trial and last observation carried forward was used if the patient dropped out of the study early.

Arm/Group Title	Insulin Lispro Mid Mixture	Insulin Glargine
Arm/Group Description	Insulin lispro mid mixture (MM) up to three times a day (TID)	Insulin glargine daily with insulin lispro at mealtime (up to 3 injections) as needed.
Measure Participants	239	240
Endpoint Hypoglycemic Episodes	87 36.4%	106 44.2%
Overall Hypoglycemic Episodes	178 74.5%	179 74.6%
Endpoint Nocturnal Hypoglycemic Episodes	30 12.6%	37 15.4%
Overall Nocturnal Hypoglycemic Episodes	112 46.9%	112 46.7%
Endpoint Non-Nocturnal Hypoglycemic Episodes	71 29.7%	98 40.8%
Overall Non-Nocturnal Hypoglycemic Episodes	175 73.2%	179 74.6%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.094
	Comments	P-value for Endpoint Hypoglycemic Episodes.
	Method	Fisher Exact
	Comments

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	1.00
	Comments	P-value for Overall Hypoglycemic Episodes.
	Method	Fisher Exact
	Comments

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.430
	Comments	P-value for Endpoint Nocturnal Hypoglycemic Episodes.
	Method	Fisher Exact
	Comments

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	1.00
	Comments	P-value for Overall Nocturnal Hypoglycemic Episodes.
	Method	Fisher Exact
	Comments

Statistical Analysis 5

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.013
	Comments	P-value for Endpoint Non-Nocturnal Hypoglycemic Episodes.
	Method	Fisher Exact
	Comments

Statistical Analysis 6

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.756
	Comments	P-value for Overall Non-Nocturnal Hypoglycemic Episodes.
	Method	Fisher Exact
	Comments

7. Secondary Outcome

Title	30-Day Adjusted Rates of Self-Reported Hypoglycemic Episodes (Including Nocturnal and Non-Nocturnal)
Description	Hypoglycemic episode defined: any time patient felt that he/she was experiencing a sign or symptom associated with hypoglycemia, or had old Roche blood glucose level <70 mg/dL even if not associated with signs, symptoms, or treatment consistent with current guidelines. Nocturnal hypoglycemia defined: any hypoglycemic event that occurred between bedtime and waking. Non-nocturnal hypoglycemia defined: any hypoglycemic event that occurred between waking and bedtime. Overall episodes: those that occurred at any time during the post-randomization visits. Endpoint: last visit interval based on LOCF.
Time Frame	Baseline to 36 Weeks

Outcome Measure Data

Analysis Population Description

Number of randomized patients with baseline and at least one post-baseline value. Intent to treat population. The last visit of trial interval was used to calculate hypoglycemic episodes at Endpoint if patient completes trial and last observation carried forward was used if the patient dropped out of the study early.

Arm/Group Title	Insulin Lispro Mid Mixture	Insulin Glargine
Arm/Group Description	Insulin lispro mid mixture (MM) up to three times a day (TID)	Insulin glargine daily with insulin lispro at mealtime (up to 3 injections) as needed.
Measure Participants	239	240
Endpoint Hypoglycemic Rate	1.57 (2.98)	2.19 (3.60)
Overall Hypoglycemic Rate	1.88 (2.44)	2.21 (2.87)
Endpoint Nocturnal Hypoglycemic Rate	0.26 (0.79)	0.32 (0.84)
Overall Nocturnal Hypoglycemic Rate	0.26 (0.43)	0.28 (0.50)
Endpoint Non-Nocturnal Hypoglycemic Rate	1.31 (2.80)	1.87 (3.29)
Overall Non-Nocturnal Hypoglycemic Rate	1.63 (2.25)	1.93 (2.71)

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.022
	Comments	P-value for Endpoint Hypoglycemic Rate.
	Method	ANOVA
	Comments	ANOVA Model: Rank of Variable = Treatment + Sufonylurea Stratum + Country + Baseline HbA1c Stratum.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.218
	Comments	P-value for Overall Hypoglycemic Rate.
	Method	ANOVA
	Comments	ANOVA Model: Rank of Variable = Treatment + Sufonylurea Stratum + Country + Baseline HbA1c Stratum.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.311
	Comments	P-value for Endpoint Nocturnal Hypoglycemic Rate.
	Method	ANOVA
	Comments	ANOVA Model: Rank of Variable = Treatment + Sufonylurea Stratum + Country + Baseline HbA1c Stratum.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.615
	Comments	P-value for Overall Nocturnal Hypoglycemic Rate.
	Method	ANOVA
	Comments	ANOVA Model: Rank of Variable = Treatment + Sufonylurea Stratum + Country + Baseline HbA1c Stratum.

Statistical Analysis 5

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.018
	Comments	P-value for Endpoint Non-Nocturnal Hypoglycemic Rate.
	Method	ANOVA
	Comments	ANOVA Model: Rank of Variable = Treatment + Sufonylurea Stratum + Country + Baseline HbA1c Stratum.

Statistical Analysis 6

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.255
	Comments	P-value for Overall Non-Nocturnal Hypoglycemic Rate.
	Method	ANOVA
	Comments	ANOVA Model: Rank of Variable = Treatment + Sufonylurea Stratum + Country + Baseline HbA1c Stratum.

8. Secondary Outcome

Title	Number of Patients With at Least One Severe Hypoglycemia Episode
Description	Severe hypoglycemia was defined as hypoglycemic event that meets at least one of the following criteria: not capable of treating self and blood glucose <2.8 millimoles per liter (mmol/L); not capable of treating self, blood glucose is missing and prompt recovery after oral carbohydrate or glucagon or intravenous glucose; hypoglycemic event outcome was coma, hopitalization, emergency room visit, or automobile accident. The overall category is a severe hypoglycemic event that occurred at any time during the post-randomization visits. Endpoint: last visit interval based on LOCF.
Time Frame	Baseline to 36 Weeks

Outcome Measure Data

Analysis Population Description

Number of randomized patients with baseline and at least one post-baseline value. Intent to treat population. Last observation carried forward.

Arm/Group Title	Insulin Lispro Mid Mixture	Insulin Glargine
Arm/Group Description	Insulin lispro mid mixture (MM) up to three times a day (TID)	Insulin glargine daily with insulin lispro at mealtime (up to 3 injections) as needed.
Measure Participants	239	240
Endpoint Severe Hypoglycemic Episodes	1 0.4%	1 0.4%
Overall Severe Hypoglycemic Episodes	8 3.3%	5 2.1%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.416
	Comments	P-value for Overall Severe Hypoglycemic Episodes.
	Method	Fisher Exact
	Comments

9. Secondary Outcome

Title	Endpoint Insulin Dose Per Body Weight; Total, Basal, and Prandial
Description	Total daily insulin dose adjusted for body weight (Units of insulin per kilogram per day [U/kg/day]) was assessed. Basal insulin is the amount of insulin required to manage normal daily blood glucose fluctuations. Prandial insulin is taken at meal time. Insulin glargine is a basal insulin and insulin lispro is a prandial insulin. Insulin lispro mid-mix is a 50/50 mixture of a basal insulin and insulin lispro. Endpoint: last visit interval based on LOCF.
Time Frame	36 Weeks

Outcome Measure Data

Analysis Population Description

Number of randomized patients with baseline and at least one post-baseline value. Intent to treat population. Last observation carried forward.

Arm/Group Title	Insulin Lispro Mid Mixture	Insulin Glargine
Arm/Group Description	Insulin lispro mid mixture (MM) up to three times a day (TID)	Insulin glargine daily with insulin lispro at mealtime (up to 3 injections) as needed.
Measure Participants	239	240
Daily Basal (n=239, n=240)	0.27 (0.07)	0.35 (0.07)
Daily Prandial (n=239, n=130)	0.36 (0.08)	0.29 (0.08)
Daily Total (n=239, n=240)	0.57 (0.11)	0.51 (0.11)

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	P-value for Daily Basal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Country + Baseline HbA1c stratum + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-0.08
	Confidence Interval	() 95%; -0.12 to -0.05
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	P-value for Daily Prandial.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Country + Baseline HbA1c stratum + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.07
	Confidence Interval	() 95%; 0.04 to 0.11
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.017
	Comments	P-value for Daily Total.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Country + Baseline HbA1c stratum + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.07
	Confidence Interval	() 95%; 0.01 to 0.12
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

10. Secondary Outcome

Title	Endpoint Insulin Dose; Total, Basal, and Prandial
Description	Total daily insulin dose (Units of insulin per day [U/day]) was assessed. Basal insulin is the amount of insulin required to manage normal daily blood glucose fluctuations. Prandial insulin is taken at meal time. Insulin glargine is a basal insulin and insulin lispro is a prandial insulin. Insulin lispro mid-mix is a 50/50 mixture of a basal insulin and insulin lispro. Endpoint: last visit interval based on LOCF.
Time Frame	36 Weeks

Outcome Measure Data

Analysis Population Description

Number of randomized patients with baseline and at least one post-baseline value. Intent to treat population. Last observation carried forward.

Arm/Group Title	Insulin Lispro Mid Mixture	Insulin Glargine
Arm/Group Description	Insulin lispro mid mixture (MM) up to three times a day (TID)	Insulin glargine daily with insulin lispro at mealtime (up to 3 injections) as needed.
Measure Participants	239	240
Daily Basal (n=239, n=240)	19.90 (5.46)	26.86 (5.37)
Daily Prandial (n=239, n=130)	27.73 (6.23)	21.68 (6.19)
Daily Total (n=239, n=240)	43.44 (8.65)	38.25 (8.51)

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	P-value for Daily Basal.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Country + Baseline HbA1c stratum + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-6.96
	Confidence Interval	() 95%; -9.65 to -4.26
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	P-value for Daily Prandial.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Country + Baseline HbA1c stratum + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	6.05
	Confidence Interval	() 95%; 3.39 to 8.71
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.017
	Comments	P-value for Daily Total.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Country + Baseline HbA1c stratum + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	5.20
	Confidence Interval	() 95%; 0.93 to 9.46
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

11. Secondary Outcome

Title	Number of Insulin Injections Per Day
Time Frame	Weeks 12, 24, 30, 36

Outcome Measure Data

Analysis Population Description

Number of randomized patients with baseline and at least one post-baseline value. Intent to treat population. If the patient completed the trial, the endpoint was 36 weeks (Week 36), but if the patient dropped out of the study early, the last observation carried forward (LOCF) was considered as the endpoint (Endpoint).

Arm/Group Title	Insulin Lispro Mid Mixture	Insulin Glargine
Arm/Group Description	Insulin lispro mid mixture (MM) up to three times a day (TID)	Insulin glargine daily with insulin lispro at mealtime (up to 3 injections) as needed.
Measure Participants	239	240
Week 12	1.33 (0.19)	1.00 (0.18)
Week 24	1.97 (0.28)	1.76 (0.27)
Week 30	2.05 (0.30)	1.81 (0.30)
Week 36	2.03 (0.30)	1.80 (0.29)
Endpoint	1.98 (0.30)	1.79 (0.29)

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	P-value for Week 12.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline HbA1c stratum + Sulfonylurea stratum + Country.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.33
	Confidence Interval	() 95%; 0.24 to 0.42
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.003
	Comments	P-value for Week 24.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline HbA1c stratum + Sulfonylurea stratum + Country.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.22
	Confidence Interval	() 95%; 0.07 to 0.36
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.002
	Comments	P-value for Week 30.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Baseline HbA1c stratum + Country + Sulfonylurea stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.24
	Confidence Interval	() 95%; 0.09 to 0.40
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.005
	Comments	P-value for Week 36.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline HbA1c stratum + Sulfonylurea stratum + Country.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.22
	Confidence Interval	() 95%; 0.07 to 0.38
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Statistical Analysis 5

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.011
	Comments	P-value for Endpoint. (last observation carried forward)
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline HbA1c stratum + Sulfonylurea stratum + Country.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	0.19
	Confidence Interval	() 95%; 0.04 to 0.34
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

12. Secondary Outcome

Title	Change From Baseline in Absolute Body Weight at 36 Week Endpoint
Description	Change in body weight was calculated as weight at endpoint (last observation carried forward) minus weight at baseline.
Time Frame	Baseline, 36 Weeks

Outcome Measure Data

Analysis Population Description

Number of randomized patients with baseline and at least one post-baseline value. Intent to treat population. Last observation carried forward.

Arm/Group Title	Insulin Lispro Mid Mixture	Insulin Glargine
Arm/Group Description	Insulin lispro mid mixture (MM) up to three times a day (TID)	Insulin glargine daily with insulin lispro at mealtime (up to 3 injections) as needed.
Measure Participants	239	240
Baseline	76.71 (5.09)	77.00 (5.01)
Change from Baseline	3.09 (1.44)	3.19 (1.42)

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Insulin Lispro Mid Mixture, Insulin Glargine
	Comments
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.803
	Comments	P-value for Change from Baseline.
	Method	ANCOVA
	Comments	ANCOVA Model: Variable=Treatment + Baseline + Sulfonylurea stratum + Country + Baseline HbA1c stratum.
Method of Estimation	Estimation Parameter	Mean Difference (Net)
	Estimated Value	-0.09
	Confidence Interval	() 95%; -0.80 to 0.62
	Parameter Dispersion	Type: Value:
	Estimation Comments	The two-sided 95% confidence interval of the Least Squares Mean difference between the two treatments is for Insulin Lispro Mid Mix minus Insulin Glargine.

Adverse Events

Arm/Group Title	Insulin Lispro Mid Mixture		Insulin Glargine
Arm/Group Description	Insulin lispro mid mixture (MM) up to three times a day (TID)		Insulin glargine daily with insulin lispro at mealtime (up to 3 injections) as needed.
All Cause Mortality
	Insulin Lispro Mid Mixture		Insulin Glargine
	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events
Total	/ (NaN)		/ (NaN)
Serious Adverse Events
	Insulin Lispro Mid Mixture		Insulin Glargine
	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events
Total	11/239 (4.6%)		11/240 (4.6%)
Blood and lymphatic system disorders
Anaemia	0/239 (0%)	0	1/240 (0.4%)	1
Cardiac disorders
Angina pectoris	0/239 (0%)	0	1/240 (0.4%)	1
Atrial fibrillation	1/239 (0.4%)	1	0/240 (0%)	0
Cardiac failure	0/239 (0%)	0	1/240 (0.4%)	1
Coronary artery disease	0/239 (0%)	0	1/240 (0.4%)	1
Myocardial infarction	1/239 (0.4%)	1	0/240 (0%)	0
Myocardial ischaemia	1/239 (0.4%)	1	0/240 (0%)	0
Eye disorders
Glaucoma	1/239 (0.4%)	1	0/240 (0%)	0
Gastrointestinal disorders
Abdominal pain lower	0/239 (0%)	0	1/240 (0.4%)	1
Gastritis	0/239 (0%)	0	1/240 (0.4%)	1
Intestinal obstruction	0/239 (0%)	0	1/240 (0.4%)	1
General disorders
Death	0/239 (0%)	0	1/240 (0.4%)	1
Pyrexia	0/239 (0%)	0	1/240 (0.4%)	1
Infections and infestations
Pneumonia	0/239 (0%)	0	1/240 (0.4%)	1
Injury, poisoning and procedural complications
Acetabulum fracture	1/239 (0.4%)	1	0/240 (0%)	0
Fall	1/239 (0.4%)	1	0/240 (0%)	0
Ilium fracture	1/239 (0.4%)	1	0/240 (0%)	0
Pelvic fracture	1/239 (0.4%)	1	0/240 (0%)	0
Post procedural complication	1/239 (0.4%)	1	0/240 (0%)	0
Rib fracture	1/239 (0.4%)	1	0/240 (0%)	0
Road traffic accident	1/239 (0.4%)	1	0/240 (0%)	0
Metabolism and nutrition disorders
Hypoglycaemia	1/239 (0.4%)	1	0/240 (0%)	0
Hypoglycaemic seizure	1/239 (0.4%)	1	0/240 (0%)	0
Neoplasms benign, malignant and unspecified (incl cysts and polyps)
Breast cancer	0/239 (0%)	0	1/240 (0.4%)	1
Non-small cell lung cancer	0/239 (0%)	0	1/240 (0.4%)	1
Renal cell carcinoma	1/239 (0.4%)	1	0/240 (0%)	0
Nervous system disorders
Convulsion	1/239 (0.4%)	1	0/240 (0%)	0
Headache	0/239 (0%)	0	1/240 (0.4%)	1
Respiratory, thoracic and mediastinal disorders
Acute pulmonary oedema	0/239 (0%)	0	1/240 (0.4%)	1
Haemothorax	1/239 (0.4%)	1	0/240 (0%)	0
Pulmonary embolism	1/239 (0.4%)	1	0/240 (0%)	0
Pulmonary oedema	1/239 (0.4%)	1	1/240 (0.4%)	1
Vascular disorders
Angiopathy	0/239 (0%)	0	1/240 (0.4%)	1
Arterial disorder	1/239 (0.4%)	1	0/240 (0%)	0
Extremity necrosis	1/239 (0.4%)	1	0/240 (0%)	0
Phlebitis	1/239 (0.4%)	1	0/240 (0%)	0
Vasculitis	0/239 (0%)	0	1/240 (0.4%)	1
Other (Not Including Serious) Adverse Events
	Insulin Lispro Mid Mixture		Insulin Glargine
	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events
Total	91/239 (38.1%)		92/240 (38.3%)
Eye disorders
Cataract	0/239 (0%)	0	3/240 (1.3%)	4
Diabetic retinopathy	1/239 (0.4%)	1	4/240 (1.7%)	4
Gastrointestinal disorders
Abdominal pain upper	3/239 (1.3%)	3	1/240 (0.4%)	1
Constipation	3/239 (1.3%)	4	0/240 (0%)	0
Diarrhoea	7/239 (2.9%)	9	8/240 (3.3%)	9
Nausea	3/239 (1.3%)	4	3/240 (1.3%)	3
Vomiting	6/239 (2.5%)	6	2/240 (0.8%)	2
General disorders
Fatigue	4/239 (1.7%)	5	2/240 (0.8%)	2
Oedema peripheral	3/239 (1.3%)	3	2/240 (0.8%)	2
Pain	4/239 (1.7%)	7	0/240 (0%)	0
Pyrexia	3/239 (1.3%)	5	1/240 (0.4%)	1
Infections and infestations
Bronchitis	6/239 (2.5%)	6	12/240 (5%)	12
Gastroenteritis	3/239 (1.3%)	4	3/240 (1.3%)	3
Influenza	3/239 (1.3%)	5	3/240 (1.3%)	3
Nasopharyngitis	20/239 (8.4%)	31	21/240 (8.8%)	22
Upper respiratory tract infection	4/239 (1.7%)	7	6/240 (2.5%)	6
Urinary tract infection	5/239 (2.1%)	5	4/240 (1.7%)	4
Metabolism and nutrition disorders
Hyperlipidaemia	3/239 (1.3%)	3	1/240 (0.4%)	1
Hypoglycaemia	3/239 (1.3%)	3	4/240 (1.7%)	4
Musculoskeletal and connective tissue disorders
Arthralgia	2/239 (0.8%)	2	6/240 (2.5%)	6
Back pain	5/239 (2.1%)	5	3/240 (1.3%)	3
Musculoskeletal pain	3/239 (1.3%)	3	2/240 (0.8%)	3
Myalgia	3/239 (1.3%)	3	0/240 (0%)	0
Pain in extremity	5/239 (2.1%)	5	1/240 (0.4%)	1
Nervous system disorders
Dizziness	3/239 (1.3%)	4	2/240 (0.8%)	3
Headache	4/239 (1.7%)	6	1/240 (0.4%)	1
Psychiatric disorders
Insomnia	4/239 (1.7%)	4	1/240 (0.4%)	1
Respiratory, thoracic and mediastinal disorders
Cough	4/239 (1.7%)	4	3/240 (1.3%)	3
Nasal congestion	3/239 (1.3%)	5	1/240 (0.4%)	1
Pharyngolaryngeal pain	3/239 (1.3%)	4	7/240 (2.9%)	8
Skin and subcutaneous tissue disorders
Rash	3/239 (1.3%)	3	1/240 (0.4%)	1
Skin ulcer	3/239 (1.3%)	3	0/240 (0%)	0
Vascular disorders
Hypertension	9/239 (3.8%)	9	4/240 (1.7%)	4

Limitations/Caveats

[Not Specified]

More Information

Certain Agreements

Principal Investigators are NOT employed by the organization sponsoring the study.

The only disclosure restriction on the PI is that the sponsor can review results communications prior to public release and can embargo communications regarding trial results for a period that is more than 60 days but less than or equal to 180 days. The sponsor cannot require changes to the communication and cannot extend the embargo.