PATIENT: Promoting Adherence to Improve Effectiveness of Cardiovascular Disease Therapies

• Sponsor: Kaiser Permanente (Other)
• Overall Status: Completed
• CT.gov ID: NCT01251757
• Collaborator: Kaiser Foundation Hospitals, Center for Health Research (Other), Johns Hopkins University (Other), Agency for Healthcare Research and Quality (AHRQ) (U.S. Fed)
• Enrollment: 21,752
• Locations: 3
• Arms: 3
• Duration (Months): 24
• Patients Per Site: 7250.7
• Patients Per Site Per Month: 301.9

Study Details

Study Description

Brief Summary

The purpose of this randomized clinical trial is to determine whether two low-intensity, technology based interventions, when compared to each other and to usual care, improve adherence to selected medications that are used to treat people with cardiovascular disease (CVD) and diabetes.

Condition or Disease	Intervention/Treatment	Phase
Diabetes Mellitus; Cardiovascular Disease	Other: Interactive Voice Recognition (IVR) phone calls; Other: Educational mailings and follow-up for nonadherence	N/A

Detailed Description

The frequent failure of patients to adhere to long-term medication regimens remains the single greatest challenge for chronic-disease management. Many studies have linked medication non-adherence to treatment failure; unnecessary and dangerous intensification of therapy; and excess health care costs, hospitalizations, and deaths. Although some interventions have been shown to significantly enhance medication adherence, the strategies used are often complex, labor-intensive, and of variable effectiveness. Simple interventions designed to make small-but-significant improvements in population-based adherence may thus offer a novel, cost-effective, and easily-disseminated alternative to current approaches for enhancing adherence. The proposed PATIENT study will use health information technology (automated phone calls and access to an electronic medical record) to test two such interventions and compare them to each other and to usual care alone.

Study Design

• Study Type: Interventional
• Actual Enrollment: 21752 participants
• Allocation: Randomized
• Intervention Model: Parallel Assignment
• Masking: None (Open Label)
• Primary Purpose: Health Services Research
• Official Title: Promoting Adherence to Improve Effectiveness of Cardiovascular Disease Therapies (PATIENT)
• Study Start Date: Aug 1, 2011
• Actual Primary Completion Date: Aug 1, 2013
• Actual Study Completion Date: Aug 1, 2013

Arms and Interventions

Arm	Intervention/Treatment
No Intervention: Usual Care (UC) Participants in this arm received their usual care with no restrictions.
Active Comparator: Interactive Voice Recognition (IVR) automated phone calls	Other: Interactive Voice Recognition (IVR) phone calls; The IVR intervention consisted of automated phone calls designed to educate participants about their medications and to assist them in refilling their prescriptions. The calls fell into two basic types: simple refill reminders and "tardy" calls for those who were overdue for a refill. Calls occured monthly and were triggered by dispensing information in the electronic medical record (EMR). Call features included the ability to transfer individuals to Kaiser's automated prescription refill service as well as to care managers. Although the calls were triggered by and focused on use of ACE inhibitors, ARBs and statins, they also included reminders to use aspirin, which is known to also be effective for secondary prevention in this patient population.; Other Names: automated phone calls
Active Comparator: Enhanced IVR (IVR+) automated phone calls & Educational mailings and follow-up for nonadherence	Other: Interactive Voice Recognition (IVR) phone calls; The IVR intervention consisted of automated phone calls designed to educate participants about their medications and to assist them in refilling their prescriptions. The calls fell into two basic types: simple refill reminders and "tardy" calls for those who were overdue for a refill. Calls occured monthly and were triggered by dispensing information in the electronic medical record (EMR). Call features included the ability to transfer individuals to Kaiser's automated prescription refill service as well as to care managers. Although the calls were triggered by and focused on use of ACE inhibitors, ARBs and statins, they also included reminders to use aspirin, which is known to also be effective for secondary prevention in this patient population.; Other Names: automated phone calls ; Other: Educational mailings and follow-up for nonadherence; Participants received bimonthly educational materials by mail. In addition, patients received mailed refill reminder letters and their providers were notified electronically if the patients failed to refill in response to the automated calls. The educational mailings included personalized health information such as the participant's cholesterol and blood pressure readings, as well as tools for improving adherence such as frequently asked questions (FAQs) about their medications, a pocket-sized calendar for tracking refills with pertinent phone numbers and web site information and space for them to write their medical record number and prescription numbers.; Other Names: mail follow-up for nonadherence educational mailings

Outcome Measures

Primary Outcome Measures

1. Adherence to Statins [12 months post randomization]

   We used a modification of the Medication Possession Ratio (MPR) as our primary outcome measure. The MPR is computed as the number of days' supply of medication dispensed during a given time window divided by the time between the first dispensing in the window and the end of the window. Our modified MPR (mMPR) also accounted for medication that was on hand at the start of the window and ignored any days' supply that would extend beyond the end of the window. We used medication dispensing data from the Kaiser outpatient pharmacies to calculate a modified medication possession ratio (mMPR) for statins among the subset of randomized participants who were using these drugs. Nominally mMPR provides an estimate of the proportion of days during the follow-up period during which the participant was adherent to their prescribed medications.

2. Adherence to Angiotensin-Converting Enzyme Inhibitors (ACEIs) and Angiotensin Receptor Blockers (ARBs) [12 months post randomization]

   We used medication dispensing data from the Kaiser outpatient pharmacies to calculate a modified medication possession ratio (mMPR) for the subset of randomized participants who were using ACEIs or ARBs. Nominally mMPR provides an estimate of the proportion of days during the follow-up period during which the participant was adherent to their prescribed medications.

Secondary Outcome Measures

1. Percentage With Good (>80%) Statin Adherence [12 months post randomization]

   Binary indicator of good statin adherence, defined as an mMPR>0.80. 1=yes, 0=no.

2. Percentage With Good (>80%) ACEI/ARB Adherence [12 months post randomization]

   Binary indicator of good ACEI/ARB adherence, defined as an mMPR>0.80. 1=yes, 0=no.

3. Systolic Blood Pressure (SBP) [12-months post randomization]

   Mean of last 5 SBP measurements captured in the electronic medical record for the 12 months post randomization.

4. Percentage With Good (<140/90 mmHg) Blood Pressure Control [12 months post randomization]

   Using the mean of last 5 available blood pressure measurements post randomization, we defined BP control as a means systolic BP <140 mmHg and a mean diastolic BP < 90 mmHg.

5. Post Intervention Low Density Lipoprotein (LDL) Level [12 months post randomization]

   We used the latest LDL (fasting or nonfasting) available during 12 months post randomization. no missing data were imputed.

6. Percentage With Good (<=100mg/dL) Low Density Lipoprotein (LDL) Control [12 months post randomization]

   Using the last LDL measurement (fasting or nonfasting) available in the EMR post randomization, we defined good control as an LDL level <= 100 mg/dL.

Eligibility Criteria

Criteria

• Ages Eligible for Study: 40 Years and Older
• Sexes Eligible for Study: All
• Accepts Healthy Volunteers: No

Inclusion Criteria:

• Aged 40 years or older as of time of randomization.

• Flagged in KP's databases as having either diabetes or atherosclerotic cardiovascular disease(defined as coronary artery disease, peripheral vascular disease, or a history of atherosclerotic stroke) at the time of randomization

• At least one dispensing of an ACEI, ARB, or statin from a Kaiser Permanente (KP) outpatient pharmacy during the baseline year.

• Suboptimal adherence ((MPR<0.9) to either statins or ACEI/ARBs during the baseline year

• Continuous membership in KP for the 12 months prior to randomization.

• Qualified for an intervention call at the time of randomization.

Exclusion Criteria:

• Evidence in the electronic medical record (EMR) of allergy or intolerance to statins or ACE inhibitors/ARBs

• medical conditions that would contraindicate use of statins or ACEI/ARBs

• Absence of either a phone number or mailing address in the EMR

• for Kaiser Permanente Hawaii, clinics whose patients tend to fill prescriptions primarily at non-KP pharmacies

• on Kaiser Permanente's "do not contact" list or in other research studies that could add undue burden

Contacts and Locations

Locations

	Site	City	State	Country	Postal Code
1	Center for Health Research, Kaiser Permanente Southeast	Atlanta	Georgia	United States	30305
2	Center for Health Research, Kaiser Permanente Hawaii	Honolulu	Hawaii	United States	96817
3	Center for Health Research, Kaiser Permanente Northwest	Portland	Oregon	United States	97227

Sponsors and Collaborators

• Kaiser Permanente
• Kaiser Foundation Hospitals, Center for Health Research
• Johns Hopkins University
• Agency for Healthcare Research and Quality (AHRQ)

Investigators

• Principal Investigator: William M Vollmer, PhD, Center for Health Research, Kaiser Permanente Northwest

Study Documents (Full-Text)

None provided.

More Information

Publications

None provided.

• Responsible Party: Kaiser Permanente
• ClinicalTrials.gov Identifier: NCT01251757
• Other Study ID Numbers: R01HS019341-01
• First Posted: Dec 2, 2010
• Last Update Posted: Mar 13, 2017
• Last Verified: Feb 1, 2017

• Individual Participant Data (IPD) Sharing Statement: No
• Plan to Share IPD: No
• Keywords provided by Kaiser Permanente: diabetes; cardiovascular disease; randomized clinical trial; medication adherence; health information technology; adult; telephone calls; IVR
• Additional relevant MeSH terms: Cardiovascular Diseases

Study Results

Participant Flow

Arm/Group Title	Usual Care (UC)	Interactive Voice Recognition (IVR)	Enhanced IVR (IVR+)
Arm/Group Description	Participants in this arm had full access to all care they were normally entitled to as part of usual care	In addition to their usual care, participants in the Interactive Voice Recognition (IVR) arm received automated phone calls, triggered by dispensing events in the electronic medical record (EMR), to educate patients about their medications and assist them in refilling their prescriptions. The calls fell into two basic types: simple refill reminders and "tardy" calls for those who were overdue for a refill. Calls occured monthly and were triggered by dispensing information in the EMR. Call features included the ability to transfer individuals to Kaiser's automated prescription refill service as well as to care managers. Although the calls were triggered by and focused on use of ACE inhibitors, ARBs and statins, they also included reminders to use aspirin, which is known to also be effective for secondary prevention in this patient population.	Participants in the IVR+ arm received all components of the IVR intervention and in addition were mailed educational materials bimonthly during the intervention. In addition, both IVR+ participants and their primary care providers received mailed notifications when they did not fill their medications in response to the automated calls. The educational mailings included personalized health information such as the participant's cholesterol and blood pressure readings, as well as tools for improving adherence such as FAQs about their medications, a pocket-sized calendar for tracking refills with pertinent phone numbers and web site information and space for them to write their medical record number and prescription numbers.
Period Title: Overall Study
STARTED	7255	7247	7250
COMPLETED	7255	7247	7250
NOT COMPLETED	0	0	0

Baseline Characteristics

Arm/Group Title	Usual Care (UC)	Interactive Voice Recognition (IVR)	Enhanced IVR (IVR+)	Total
Arm/Group Description	Participants in this arm had full access to all care they were normally entitled to as part of usual care	In addition to their usual care, participants in the Interactive Voice Recognition (IVR) arm received automated phone calls, triggered by dispensing events in the electronic medical record (EMR), to educate patients about their medications and assist them in refilling their prescriptions. The calls fell into two basic types: simple refill reminders and "tardy" calls for those who were overdue for a refill. Calls occured monthly and were triggered by dispensing information in the EMR. Call features included the ability to transfer individuals to Kaiser's automated prescription refill service as well as to care managers. Although the calls were triggered by and focused on use of ACE inhibitors, ARBs and statins, they also included reminders to use aspirin, which is known to also be effective for secondary prevention in this patient population.	Participants in the IVR+ arm received all components of the IVR intervention and in addition were mailed educational materials bimonthly during the intervention. In addition, both IVR+ participants and their primary care providers received mailed notifications when they did not fill their medications in response to the automated calls. The educational mailings included personalized health information such as the participant's cholesterol and blood pressure readings, as well as tools for improving adherence such as FAQs about their medications, a pocket-sized calendar for tracking refills with pertinent phone numbers and web site information and space for them to write their medical record number and prescription numbers.	Total of all reporting groups
Overall Participants	7255	7247	7250	21752
Age (years) [Mean (Standard Deviation) ]
Mean (Standard Deviation) [years]	63.6 (12.2)	63.6 (12.1)	63.5 (12.2)	63.6 (12.2)
Sex: Female, Male (Count of Participants)
Female	3432 47.3%	3370 46.5%	3415 47.1%	10217 47%
Male	3823 52.7%	3877 53.5%	3835 52.9%	11535 53%
Race (NIH/OMB) (Count of Participants)
American Indian or Alaska Native	51 0.7%	43 0.6%	44 0.6%	138 0.6%
Asian	1270 17.5%	1290 17.8%	1254 17.3%	3814 17.5%
Native Hawaiian or Other Pacific Islander	798 11%	819 11.3%	783 10.8%	2400 11%
Black or African American	1168 16.1%	1094 15.1%	1109 15.3%	3371 15.5%
White	3403 46.9%	3399 46.9%	3408 47%	10210 46.9%
More than one race	0 0%	0 0%	0 0%	0 0%
Unknown or Not Reported	565 7.8%	602 8.3%	652 9%	1819 8.4%
Region of Enrollment (participants) [Number]
United States	7255 100%	7247 100%	7250 100%	21752 100%
ever smokers (participants) [Number]
yes	3482 48%	3537 48.8%	3560 49.1%	10579 48.6%
no	3773 52%	3710 51.2%	3690 50.9%	11173 51.4%
co-morbid diabetes mellitus (participants) [Number]
yes	5666 78.1%	5703 78.7%	5626 77.6%	16995 78.1%
no	1589 21.9%	1544 21.3%	1624 22.4%	4757 21.9%
co-morbid cardiovascular disease (participants) [Number]
yes	2626 36.2%	2616 36.1%	2661 36.7%	7903 36.3%
no	4629 63.8%	4631 63.9%	4589 63.3%	13849 63.7%
target medication use (participants) [Number]
statin only	2924 40.3%	2870 39.6%	2922 40.3%	8716 40.1%
ACEI/ARB only	1770 24.4%	1790 24.7%	1820 25.1%	5380 24.7%
statin and ACEI/ARB	2561 35.3%	2587 35.7%	2508 34.6%	7656 35.2%

Outcome Measures

1. Primary Outcome

Title	Adherence to Statins
Description	We used a modification of the Medication Possession Ratio (MPR) as our primary outcome measure. The MPR is computed as the number of days' supply of medication dispensed during a given time window divided by the time between the first dispensing in the window and the end of the window. Our modified MPR (mMPR) also accounted for medication that was on hand at the start of the window and ignored any days' supply that would extend beyond the end of the window. We used medication dispensing data from the Kaiser outpatient pharmacies to calculate a modified medication possession ratio (mMPR) for statins among the subset of randomized participants who were using these drugs. Nominally mMPR provides an estimate of the proportion of days during the follow-up period during which the participant was adherent to their prescribed medications.
Time Frame	12 months post randomization

Outcome Measure Data

Analysis Population Description

All randomized participants who were taking statins at the time of randomization

Arm/Group Title	Usual Care (UC)	Interactive Voice Recognition (IVR)	Enhanced IVR (IVR+)
Arm/Group Description	statin users in UC arm	statin users in IVR arm	statin users in IVR+ arm
Measure Participants	5484	5453	5429
overall	0.55 (0.35)	0.57 (0.34)	0.58 (0.34)
baseline adherence <= 0.40	0.38 (0.37)	0.41 (0.36)	0.41 (0.36)
baseline adherence 0.4-0.75	0.61 (0.31)	0.63 (0.31)	0.65 (0.30)
baseline adherence 0.75-0.90	0.75 (0.27)	0.77 (0.27)	0.75 (0.29)

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Usual Care (UC), Interactive Voice Recognition (IVR)
	Comments	We conducted separate primary analyses of each of IVR and IVR+ versus usual care at the .025 level of significance to assure a trial-wide error rate of .05. This analysis summary is for the IVR versus UC comparison only. Also, we only report results here for the full sample, and not also separately for the 3 baseline adherence level groups.
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	<.001
	Comments
	Method	Regression, Linear
	Comments	adjusted for site, gender, age, total number of prescription medications patient is taking, comorbid diabetes/CVD, and baseline statin adherence.
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	0.022
	Confidence Interval	(2-Sided) 95%; 0.011 to 0.034
	Parameter Dispersion	Type: Value:
	Estimation Comments	Adjusted difference in adherence for IVR group versus UC group, calculated as IVR - UC.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Usual Care (UC), Enhanced IVR (IVR+)
	Comments	We conducted separate primary analyses of each of IVR and IVR+ versus usual care at the .025 level of significance to assure a trial-wide error rate of .05. This analysis summary is for the IVR+ versus UC comparison only. Also, we only report results here for the full sample, and not also separately for the 3 baseline adherence level groups.
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	<.001
	Comments
	Method	Regression, Linear
	Comments	adjusted for site, gender, age, total number of prescription medications participant is taking, comorbid diabetes/CVD, and baseline statin adherence.
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	0.030
	Confidence Interval	(2-Sided) 95%; 0.019 to 0.042
	Parameter Dispersion	Type: Value:
	Estimation Comments	Adjusted difference in adherence for IVR+ group versus UC group, calculated as IVR+ - UC.

2. Primary Outcome

Title	Adherence to Angiotensin-Converting Enzyme Inhibitors (ACEIs) and Angiotensin Receptor Blockers (ARBs)
Description	We used medication dispensing data from the Kaiser outpatient pharmacies to calculate a modified medication possession ratio (mMPR) for the subset of randomized participants who were using ACEIs or ARBs. Nominally mMPR provides an estimate of the proportion of days during the follow-up period during which the participant was adherent to their prescribed medications.
Time Frame	12 months post randomization

Outcome Measure Data

Analysis Population Description

All randomized participants who were taking an ACEI or an ARB at the time of randomization

Arm/Group Title	Usual Care (UC)	Interactive Voice Recognition (IVR)	Enhanced IVR (IVR+)
Arm/Group Description	ACEI/ARB users in UC arm	ACEI/ARB users in IVR arm	ACEI/ARB users in IVR+ arm
Measure Participants	4330	4370	4323
overall	0.57 (0.36)	0.59 (0.35)	0.61 (0.34)
baseline adherence <=0.50	0.45 (0.37)	0.47 (0.37)	0.50 (0.36)
baseline adherence 0.50-0.75	0.69 (0.30)	0.70 (0.29)	0.71 (0.29)
baseline adherence 0.75-0.90	0.76 (0.26)	0.76 (0.28)	0.77 (0.29)

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Usual Care (UC), Interactive Voice Recognition (IVR)
	Comments	We conducted separate primary analyses of each of IVR and IVR+ versus usual care at the .025 level of significance to assure a trial-wide error rate of .05. This analysis summary is for the IVR versus UC comparison only. Also, we only report results here for the full sample, and not also separately for the 3 baseline adherence level groups.
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.022
	Comments
	Method	Regression, Linear
	Comments	Adjusted for site, gender, age, total number of prescription medications participant was taking, comorbid diabetes/CVD, baseline ACEI/ARB adherence.
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	0.016
	Confidence Interval	(2-Sided) 95%; 0.002 to 0.029
	Parameter Dispersion	Type: Value:
	Estimation Comments	Adjusted difference in adherence for IVR+ group versus UC group, calculated as IVR+ - UC.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Usual Care (UC), Enhanced IVR (IVR+)
	Comments	We conducted separate primary analyses of each of IVR and IVR+ versus usual care at the .025 level of significance to assure a trial-wide error rate of .05. This analysis summary is for the IVR+ versus UC comparison only. Also, we only report results here for the full sample, and not also separately for the 3 baseline adherence level groups.
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	<.001
	Comments
	Method	Regression, Linear
	Comments	Adjusted for site, gender, age, total number of prescription medications participant was taking, comorbid diabetes/CVD, baseline ACEI/ARB adherence.
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	0.037
	Confidence Interval	(2-Sided) 95%; 0.023 to 0.050
	Parameter Dispersion	Type: Value:
	Estimation Comments	Adjusted difference in adherence for IVR+ group versus UC group, calculated as IVR+ - UC.

3. Secondary Outcome

Title	Percentage With Good (>80%) Statin Adherence
Description	Binary indicator of good statin adherence, defined as an mMPR>0.80. 1=yes, 0=no.
Time Frame	12 months post randomization

Outcome Measure Data

Analysis Population Description

All randomized participants who were taking statins at the time of randomization.

Arm/Group Title	Usual Care (UC)	Interactive Voice Recognition (IVR)	Enhanced IVR (IVR+)
Arm/Group Description	statin users in UC arm	statin users in IVR arm	statin users in IVR+ arm
Measure Participants	5484	5453	5429
Overall	32.9	35.9	35.8
baseline adherence <=0.40	21.5	22.6	21.8
baseline adherence 0.4-0.75	36.2	39.4	40.1
baseline adherence 0.75-0.90	58.1	63.5	62.6

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Usual Care (UC), Interactive Voice Recognition (IVR)
	Comments	We conducted separate primary analyses of each of IVR and IVR+ versus usual care at the .025 level of significance to assure a trial-wide error rate of .05. This analysis summary is for the IVR versus UC comparison only. Also, we only report results here for the full sample, and not also separately for the 3 baseline adherence level groups.
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.002
	Comments	Adjusted for site, gender, age, total number of prescription medications patient is taking, comorbid diabetes/CVD, and baseline statin adherence.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.14
	Confidence Interval	(2-Sided) 95%; 1.05 to 1.24
	Parameter Dispersion	Type: Value:
	Estimation Comments	Adjusted odds ratio for good adherence in IVR group versus UC group, calculated as IVR - UC.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Usual Care (UC), Enhanced IVR (IVR+)
	Comments	We conducted separate primary analyses of each of IVR and IVR+ versus usual care at the .025 level of significance to assure a trial-wide error rate of .05. This analysis summary is for the IVR+ versus UC comparison only. Also, we only report results here for the full sample, and not also separately for the 3 baseline adherence level groups.
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	Adjusted for site, gender, age, total number of prescription medications patient is taking, comorbid diabetes/CVD, and baseline statin adherence.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.16
	Confidence Interval	(2-Sided) 95%; 1.06 to 1.26
	Parameter Dispersion	Type: Value:
	Estimation Comments	Adjusted odds ratio for good adherence in IVR+ group versus UC group, calculated as IVR+ - UC.

4. Secondary Outcome

Title	Percentage With Good (>80%) ACEI/ARB Adherence
Description	Binary indicator of good ACEI/ARB adherence, defined as an mMPR>0.80. 1=yes, 0=no.
Time Frame	12 months post randomization

Outcome Measure Data

Analysis Population Description

All randomized participants who were taking an ACEI or an ARB at the time of randomization

Arm/Group Title	Usual Care (UC)	Interactive Voice Recognition (IVR)	Enhanced IVR (IVR+)
Arm/Group Description	ACEI/ARB users in UC arm	ACEI/ARB users in IVR arm	ACEI/ARB users in IVR+ arm
Measure Participants	4330	4370	4323
Overall	37.4	40.3	41.6
Baseline adherence <=0.50	24.5	27.0	28.2
Baseline adherence 0.50-0.75	48.4	51.1	52.0
Baseline adherence 0.75-0-.90	59.7	62.0	66.8

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Usual Care (UC), Interactive Voice Recognition (IVR)
	Comments	We conducted separate primary analyses of each of IVR and IVR+ versus usual care at the .025 level of significance to assure a trial-wide error rate of .05. This analysis summary is for the IVR versus UC comparison only. Also, we only report results here for the full sample, and not also separately for the 3 baseline adherence level groups.
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	0.014
	Comments	adjusted for site, gender, age, total number of prescription medications participant was taking, comorbid diabetes/CVD, baseline ACEI/ARB adherence
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.12
	Confidence Interval	(2-Sided) 95%; 1.02 to 1.23
	Parameter Dispersion	Type: Value:
	Estimation Comments	Adjusted odds ratio for good adherence in IVR group versus UC group, calculated as IVR - UC.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Usual Care (UC), Enhanced IVR (IVR+)
	Comments	We conducted separate primary analyses of each of IVR and IVR+ versus usual care at the .025 level of significance to assure a trial-wide error rate of .05. This analysis summary is for the IVR+ versus UC comparison only. Also, we only report results here for the full sample, and not also separately for the 3 baseline adherence level groups.
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	Adjusted for site, gender, age, total number of prescription medications patient is taking, comorbid diabetes/CVD, and baseline statin adherence.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.21
	Confidence Interval	(2-Sided) 95%; 1.10 to 1.32
	Parameter Dispersion	Type: Value:
	Estimation Comments	Adjusted odds ratio for good adherence in IVR group versus UC group, calculated as IVR - UC.

5. Secondary Outcome

Title	Systolic Blood Pressure (SBP)
Description	Mean of last 5 SBP measurements captured in the electronic medical record for the 12 months post randomization.
Time Frame	12-months post randomization

Outcome Measure Data

Analysis Population Description

The analysis sample was restricted to ACEI/ARB users with at least one post intervention SBP measurement recorded in the EMR. We did not impute any missing data.

Arm/Group Title	Usual Care (UC)	Interactive Voice Recognition (IVR)	Enhanced IVR (IVR+)
Arm/Group Description	ACEI/ARB users in UC arm	ACEI/ARB users in IVR arm .	ACEI/ARB users in IVR+ arm
Measure Participants	3980	3980	3971
Overall	129.2 (13.1)	128.6 (13.2)	129.0 (13.2)
Baseline SBP<=130mmHg	123.5 (10.7)	122.8 (11.1)	123.4 (11.1)
Baseline SBP 130-140mmHg	132.8 (10.7)	132.2 (10.1)	132.7 (10.1)
Baseline SBP>140mmHg	140.6 (13.3)	140.4 (13.2)	140.9 (13.2)

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Usual Care (UC), Interactive Voice Recognition (IVR)
	Comments	We conducted separate primary analyses of each of IVR and IVR+ versus usual care at the .025 level of significance to assure a trial-wide error rate of .05. This analysis summary is for the IVR versus UC comparison only. Also, we only report results here for the full sample, and not also separately for the 3 baseline SBP level groups.
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	.041
	Comments	As noted above, the a priori threshold for statistical significance for this test was .025.
	Method	Regression, Linear
	Comments	Adjusted for site, gender, age, total number of prescription medications patient is taking, comorbid diabetes/CVD, and baseline SBP group.
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	-0.5
	Confidence Interval	(2-Sided) 95%; -1.0 to 0.0
	Parameter Dispersion	Type: Value:
	Estimation Comments

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Usual Care (UC), Enhanced IVR (IVR+)
	Comments	We conducted separate primary analyses of each of IVR and IVR+ versus usual care at the .025 level of significance to assure a trial-wide error rate of .05. This analysis summary is for the IVR+ versus UC comparison only. Also, we only report results here for the full sample, and not also separately for the 3 baseline SBP level groups.
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	.93
	Comments	As noted above, the a priori threshold for statistical significance for this test was .025.
	Method	Regression, Linear
	Comments	Adjusted for site, gender, age, total number of prescription medications participant is taking, comorbid diabetes/CVD, and baseline SBP group.
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	0.0
	Confidence Interval	(2-Sided) 95%; -.5 to .5
	Parameter Dispersion	Type: Value:
	Estimation Comments

6. Secondary Outcome

Title	Percentage With Good (<140/90 mmHg) Blood Pressure Control
Description	Using the mean of last 5 available blood pressure measurements post randomization, we defined BP control as a means systolic BP <140 mmHg and a mean diastolic BP < 90 mmHg.
Time Frame	12 months post randomization

Outcome Measure Data

Analysis Population Description

All randomized participants who were taking an ACEI or an ARB at the time of randomization and who had at least one post randomization BP recorded in the EMR. Missing data were not imputed.

Arm/Group Title	Usual Care (UC)	Interactive Voice Recognition (IVR)	Enhanced IVR (IVR+)
Arm/Group Description	ACEI/ARB users in UC arm	ACEI/ARB users in IVR arm .	ACEI/ARB users in IVR+ arm
Measure Participants	3980	3980	3971
Overall	81.1	82.1	81.3
Baseline SBP<=130 mmHg	93.7	94.0	92.9
Baseline SBP 130-140 mmHg	78.4	78.4	78.3
Baseline SBP>140 mmHg	49.2	52.4	50.2

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Usual Care (UC), Interactive Voice Recognition (IVR)
	Comments	We conducted separate primary analyses of each of IVR and IVR+ versus usual care at the .025 level of significance to assure a trial-wide error rate of .05. This analysis summary is for the IVR versus UC comparison only. Also, we only report results here for the full sample, and not also separately for the 3 baseline SBP level groups.
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	.404
	Comments	As noted above, the a priori threshold for statistical significance for this test was .025.
	Method	Regression, Logistic
	Comments	Adjusted for site, gender, age, total number of prescription medications participant is taking, comorbid diabetes/CVD, and baseline SBP group
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.05
	Confidence Interval	(2-Sided) 95%; 0.93 to 1.19
	Parameter Dispersion	Type: Value:
	Estimation Comments

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Usual Care (UC), Enhanced IVR (IVR+)
	Comments	We conducted separate primary analyses of each of IVR and IVR+ versus usual care at the .025 level of significance to assure a trial-wide error rate of .05. This analysis summary is for the IVR+ versus UC comparison only. Also, we only report results here for the full sample, and not also separately for the 3 baseline SBP level groups.
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	.54
	Comments	As noted above, the a priori threshold for statistical significance for this test was .025.
	Method	Regression, Logistic
	Comments	Adjusted for site, gender, age, total number of prescription medications participant is taking, comorbid diabetes/CVD, and baseline SBP group
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	.96
	Confidence Interval	(2-Sided) 95%; .85 to 1.09
	Parameter Dispersion	Type: Value:
	Estimation Comments

7. Secondary Outcome

Title	Post Intervention Low Density Lipoprotein (LDL) Level
Description	We used the latest LDL (fasting or nonfasting) available during 12 months post randomization. no missing data were imputed.
Time Frame	12 months post randomization

Outcome Measure Data

Analysis Population Description

All randomized participants who were taking statin at the time of randomization and who had at least one post randomization LDL measurement recorded in the EMR. Missing data were not imputed.

Arm/Group Title	Usual Care (UC)	Interactive Voice Recognition (IVR)	Enhanced IVR (IVR+)
Arm/Group Description	statin users in UC arm	statin users in IVR arm	statin users in IVR+ arm
Measure Participants	4621	4610	4545
Overall	92.4 (35.3)	91.8 (34.0)	91.3 (33.3)
Baseline LDL <=80 mg/dL	75.5 (26.1)	75.5 (25.7)	75.6 (26.1)
Baseline LDL 80-100 mg/dL	92.0 (27.8)	91.5 (25.9)	90.7 (24.9)
Baseline LDL>100 mg/dL	114.9 (37.9)	113.0 (37.9)	111.4 (36.6)

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Usual Care (UC), Interactive Voice Recognition (IVR)
	Comments	We conducted separate primary analyses of each of IVR and IVR+ versus usual care at the .025 level of significance to assure a trial-wide error rate of .05. This analysis summary is for the IVR versus UC comparison only. Also, we only report results here for the full sample, and not also separately for the 3 baseline LDL subgroups.
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	.38
	Comments	As noted above, the a priori threshold for statistical significance for this test was .025.
	Method	Regression, Linear
	Comments	Adjusted for site, gender, age, total number of prescription medications participant is taking, comorbid diabetes/CVD, and baseline LDL group.
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	-0.6
	Confidence Interval	(2-Sided) 95%; -1.8 to 0.7
	Parameter Dispersion	Type: Value:
	Estimation Comments

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Usual Care (UC), Enhanced IVR (IVR+)
	Comments	We conducted separate primary analyses of each of IVR and IVR+ versus usual care at the .025 level of significance to assure a trial-wide error rate of .05. This analysis summary is for the IVR+ versus UC comparison only. Also, we only report results here for the full sample, and not also separately for the 3 baseline LDL level subgroups.
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	.019
	Comments	As noted above, the a priori threshold for statistical significance for this test was .025.
	Method	Regression, Linear
	Comments	adjusted for site, gender, age, total number of prescription medications participant is taking, comorbid diabetes/CVD, and baseline LDL group.
Method of Estimation	Estimation Parameter	Mean Difference (Final Values)
	Estimated Value	-1.5
	Confidence Interval	(2-Sided) 95%; -2.7 to -0.2
	Parameter Dispersion	Type: Value:
	Estimation Comments	MeanLDL levels were sig lower for IVR+ participants than for UC participants. In subgroup analyses this difference was most pronounced in those individuals with baseline LDL levels above 100 mg/dL (adj diff=-3.6 mg/dL, 95%CI= (-5.9, -1.3), p=.002).

8. Secondary Outcome

Title	Percentage With Good (<=100mg/dL) Low Density Lipoprotein (LDL) Control
Description	Using the last LDL measurement (fasting or nonfasting) available in the EMR post randomization, we defined good control as an LDL level <= 100 mg/dL.
Time Frame	12 months post randomization

Outcome Measure Data

Analysis Population Description

All randomized participants who were taking a statin at the time of randomization and who had at least one post randomization LDL measurement recorded in the EMR. Missing data were not imputed.

Arm/Group Title	Usual Care (UC)	Interactive Voice Recognition (IVR)	Enhanced IVR (IVR+)
Arm/Group Description	statin users in UC arm	statin users in IVR arm	statin users in IVR+ arm
Measure Participants	4621	4610	4545
Overall	69.1	69.8	70.4
Baseline LDL <=80 mg/dL	88.4	88.0	88.1
Baseline LDL 80-100 mg/dL	73.7	74.0	74.9
Baseline LDL > 100 mg/dL	39.9	43.5	44.1

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Usual Care (UC), Interactive Voice Recognition (IVR)
	Comments	We conducted separate primary analyses of each of IVR and IVR+ versus usual care at the .025 level of significance to assure a trial-wide error rate of .05. This analysis summary is for the IVR versus UC comparison only. Also, we only report results here for the full sample, and not also separately for the 3 baseline LDL level subgroups.
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	.59
	Comments	As noted above, the a priori threshold for statistical significance for this test was .025.
	Method	Regression, Logistic
	Comments	Adjusted for site, gender, age, total number of prescription medications participant is taking, comorbid diabetes/CVD, and baseline LDL group.
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.03
	Confidence Interval	(2-Sided) 95%; 0.93 to 1.13
	Parameter Dispersion	Type: Value:
	Estimation Comments

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Usual Care (UC), Enhanced IVR (IVR+)
	Comments	We conducted separate primary analyses of each of IVR and IVR+ versus usual care at the .025 level of significance to assure a trial-wide error rate of .05. This analysis summary is for the IVR+ versus UC comparison only. Also, we only report results here for the full sample, and not also separately for the 3 baseline LDL level subgroups.
	Type of Statistical Test	Superiority or Other
	Comments
Statistical Test of Hypothesis	p-Value	.058
	Comments	As noted above, the a priori threshold for statistical significance for this test was .025.
	Method	Regression, Logistic
	Comments	Adjusted for site, gender, age, total number of prescription medications participant is taking, comorbid diabetes/CVD, and baseline LDL group.
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.10
	Confidence Interval	(2-Sided) 95%; 1.00 to 1.22
	Parameter Dispersion	Type: Value:
	Estimation Comments	Though higher for the IVR+ group, LDL control did not differ significantly between the IVR+ and UC arms. Among those with poor initial control, however, control was sig better for the IVR+ arm (OR = 1.21, 95%CI = (1.04, 1.42), p=.015).

Adverse Events

Time Frame	12 months post randomization
Arm/Group Title	Usual Care (UC)		Interactive Voice Recognition (IVR)		Enhanced IVR (IVR+)
Arm/Group Description	usual medical care		usual care plus automated phone calls		usual care plus automated phone calls plus educational mailings and mail follow-up for persistent nonadherence
All Cause Mortality
	Usual Care (UC)		Interactive Voice Recognition (IVR)		Enhanced IVR (IVR+)
	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events
Total	141/7255 (1.9%)		146/7247 (2%)		140/7250 (1.9%)
Serious Adverse Events
	Usual Care (UC)		Interactive Voice Recognition (IVR)		Enhanced IVR (IVR+)
	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events
Total	25/7255 (0.3%)		21/7247 (0.3%)		22/7250 (0.3%)
Hepatobiliary disorders
potential statin related hospitalization	1/3980 (0%)	1	2/3980 (0.1%)	2	2/3971 (0.1%)	2
Vascular disorders
potential ACEI/ARB related hospitalization	24/4330 (0.6%)	24	19/4370 (0.4%)	19	20/4323 (0.5%)	20
Other (Not Including Serious) Adverse Events
	Usual Care (UC)		Interactive Voice Recognition (IVR)		Enhanced IVR (IVR+)
	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events
Total	0/7255 (0%)		0/7247 (0%)		0/7250 (0%)

Limitations/Caveats

[Not Specified]

More Information

Certain Agreements

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

There is NOT an agreement between Principal Investigators and the Sponsor (or its agents) that restricts the PI's rights to discuss or publish trial results after the trial is completed.