. 2007;13[pt1]:313-315), the e-mail address listed for Dr Szpunar was incorrect. The correct e-mail address is: susan.szpunar@stjohn.org.

Articles

CORRECTION STATEMENT

Szpunar et al (Am J Manag Care. 2007;13[pt1]:313-315)

2007;13[suppl 3]:S80), the street address provided for Dr Liu was incorrect. The correct address is: 235 East 42nd St, New York, NY 10017.

Liu et al (Am J Manag Care. 2007;13[suppl 3]:S80)

:To examine the association between generosity of drug coverage and essential cardiovascular medication use among retired seniors.  Study Design: Retrospective analysis of the 1997 to 2000 Medicare Current Beneficiary Survey, a nationally representative survey of the Medicare population.

The study examined community-dwelling fee-for-service Medicare beneficiaries aged 65 years or older with retiree health insurance and with coronary heart disease and hyper-lipidemia (n = 1220) or congestive heart failure (n = 1147). Generosity of drug coverage was defined as the percentage of the beneficiaryâ€™s annual drug expenditures paid by the employer. Dependent variables were any statin use for the group with coronary heart disease and hyperlipi-demia and any angiotensin-converting enzyme inhibitor or angiotensin receptor blocker use in the group with congestive heart failure. Logistic regression analyses estimated the adjusted odds of essential medication use by generosity category in each disease group. We estimated the extent to which medication use would change if generosity levels moved to those under standard Medicare Part D levels.

 The overall prevalence of statin use was 64.1%, and that of angiotensin-converting enzyme inhibitor or angiotensin receptor blocker use was 50.0%. In both disease groups, retirees in the less generous drug coverage categories had significantly lower adjusted odds of use than retirees with the most generous drug benefits (ie, covering >76% of annual drug expenditures). Overall, the shift to a standard Medicare Part D structure would result in mean declines of 8.4% in statin use and 5.2% in angiotensin-converting enzyme inhibitor or angiotensin receptor blocker use. Retirees with the most generous drug coverage face twice the mean decline in drug use.  

 Retirees who already have generous drug benefits from their employers may be placed at risk for decreased utilization of effective medications due to any future scaling back of retiree drug coverage.

With a double-digit growth in pharmaceutical expenditures, many employers have increased prescription cost-sharing for retirees in recent years. The new Medicare Part D has created additional (albeit unintended) incentives for terminating or scaling back retiree drug coverage.

Retirees who have generous drug benefits from their employers may be placed at risk for decreased utilization of effective medications due to any future scaling back of retiree drug coverage. 

One third of the Medicare beneficiaries now have drug coverage through employer-sponsored health plans, and most of these are offered by their own or their spouse's former employer or union. Most of these beneficiaries receive more generous coverage under these plans than they would under the new Medicare prescription drug coverage (Part D) which began January 1, 2006.

Competitive labor markets have provided sufficient incentive in the past few decades for employers to voluntarily offer retiree health benefits with generous drug coverage to workers and their families.

However, because an accounting rule change in the early 1990s requiring employers to include retiree health benefit liabilities on their financial statements, coupled with a double-digit growth in healthcare costs (particularly for pharmaceuticals), many employers have been overhauling their retiree health and drug benefits to control, reduce, or eliminate their retiree health benefit liabilities.

With the enactment of the Medicare Prescription Drug Improvement and Modernization Act, additional (albeit unintended) incentives for terminating or scaling back retiree drug coverage have been created. First, the existence of Medicare Part D as an alternative source of drug coverage for retirees would reduce political resistance for employers considering dropping coverage altogether. In fact, the Congressional Budget Office

estimated that 17% of retirees were likely to lose their employer-sponsored drug benefits; more recent estimates from the Department of Health and Human Services suggest that it might be double that.

Second, in light of the ongoing trends of reductions in generosity of retiree drug coverage due to business, accounting, and cost factors,

employers planning to continue offering benefits may now have Medicare Part D as a concrete example of a "floor" or minimum level of generosity for retiree drug benefits. This is because the US Congress has built in financial incentives in terms of direct subsidies of up to $88 billion to employers for maintaining their retiree drug coverage. To receive a tax-free subsidy (equal to 28% of total drug costs between $250 and $5000 for each covered retiree), the employer would need to offer employment-based retiree drug coverage that is at least actuarially equivalent to the standard Medicare Part D coverage. Hence, employers planning to continue offering drug coverage could over time reduce the generosity of their benefits all the way to Medicare Part D levels and still receive the benefit of the federal subsidies.

Regardless of whether employers will drop coverage to standard Medicare Part D levels or not, recent surveys suggest that employers are scaling back prescription coverage through increased drug cost-sharing for retirees and are likely to continue doing so during the long term.

The question of how substantial increases in out-of-pocket shares for prescription expenditures could potentially affect medication use, particularly among Medicare beneficiaries with retiree health insurance, has not been studied in detail. This study takes advantage of the existing variation in retiree drug benefits to examine the potential effect of significant changes in prescription cost-sharing on use of essential chronic medications among retired seniors. To reduce the complexity in testing this relationship, we specifically examine the association between the generosity of drug coverage and use of 2 effective cardiovascular medications selected from well-established clinical guidelines,

namely, (1) statin use among retirees with coronary heart disease (CHD) and hyperlipidemia and (2) angiotensin-converting enzyme (ACE) inhibitor or angiotensin receptor blocker (ARB) use among retirees with congestive heart failure (CHF).

The data source used for the study was the 1997 to 2000 Medicare Current Beneficiary Survey (MCBS) linked with Medicare claims. The MCBS is a nationally representative survey of Medicare beneficiaries conducted under the auspices of the Centers for Medicare & Medicaid Services. It includes detailed information on beneficiaries' demographics, health and functional status, health insurance coverage, and annual use and spending for all healthcare services, including prescription drugs. The health insurance section of the survey captures the source of supplemental insurance (eg, current employer, former employer, self-purchased, Medicaid, and others) and whether these plans offer prescription drug coverage. The prescription drug use section captures drug names during the thrice-yearly interviews of respondents who are instructed to keep a medication log, save insurance receipts, and show the interviewers all of their medication containers.

The sample frame consists of community-dwelling fee-for-service Medicare beneficiaries aged 65 years or older with supplemental health insurance from their own or their spouse's former employer from 1997 to 2000. We limited the sample to fee-for-service beneficiaries because availability of Medicare claims was necessary to identify disease conditions. From the

International Classification of Diseases, Ninth Revision, Clinical Modification 

1997 to 2000 pooled cross sections, we selected 2 groups for analysis based on 1 or more inpatient or outpatient claims for (1) CHD and hyperlipidemia (n = 1220) and (2) CHF (n = 1147) (an appendix containing the 

). For the primary analysis, we excluded Medicare retirees with additional drug coverage from sources other than an employer (153 retirees from the group with CHD and hyperlipidemia and 167 retirees from the group with CHF). However, secondary analysis without excluding this group indicated that our results are not sensitive to this sample exclusion criterion, and results only for the primary sample are presented.

Our dependent variable was an indicator of the presence of any use of the essential cardiovascular medications. The indicator represented any statin use for the group with CHD and hyperlipidemia and any ACE inhibitor or ARB use in the group with CHF.

The main independent variable of interest was the generosity of drug coverage. This variable was defined as the percentage of the beneficiary's annual drug expenditures paid by the employer. We grouped beneficiaries into the following 4 generosity categories: 0%, 1% to 50%, 51% to 75%, and 76% to 100%. The 0% category primarily represents beneficiaries with no retiree drug benefits but also includes a small group with such limited drug benefits that none of their drug spending was paid by the employer.

Control variables included socioeconomic and demographic factors such as age, sex, education, income, metropolitan status, and geographic region of residence. We also controlled for smoking status and whether the beneficiary had ever visited a cardiologist. In addition, we included the following clinical characteristics: numbers of physician visits and hospitalizations for the specific study condition, presence of contraindication to the essential medication, cardiovascular risk factors likely to increase essential medication use, number of chronic conditions other than cardiovascular disease, and self-reported health status (the details are available in an appendix from the author). Survey year of observation was also used as a control.

The adjusted odds of receiving the essential cardiovascular medication by coverage generosity level were estimated using separate logistic regression analyses in each disease group. Although the logistic regression analyses controlled for differences in available demographic and clinical measures across the generosity of drug coverage categories, it is possible that there remain unmeasured factors predictive of generosity of drug coverage and of essential medication use resulting from the mechanism of drug coverage selection. We test for the presence of this bias using the Durbin-Wu-Hausman test.

To aid in the policy interpretation of the adjusted odds ratios, we estimated the extent to which essential medication use would change if generosity levels for all retirees moved from existing levels to Medicare Part D levels in 2006. This required estimating generosity levels for all retirees under standard Medicare Part D. We did this in 2 steps. First, given that existing levels of drug expenditures are from the 1997 to 2000 MCBS data and that the standard Medicare Part D benefit limits and thresholds were designed for 2006, we projected annual drug spending from each year of the 1997 to 2000 MCBS to 2006 using historical and projected per capita drug-spending data from the National Health Accounts (and assuming no Medicare Part D implementation).

These forecasts represent what one would expect retiree drug spending to be in 2006 if Medicare Part D had not been enacted. Second, we imposed the standard Medicare Part D benefit structure on these spending estimates to compute the new levels of generosity of drug coverage. The standard Medicare Part D plan in 2006 required all (except low-income) beneficiaries to pay an annual deductible of $250, 25% coinsurance for drug spending from $250 to $2250, 100% coinsurance for the next $2850 of drug expenses (ie, the "doughnut hole"), and 5% coinsurance after $3600 in out-of-pocket costs (ie, $5100 in total expenses). Hence, for example, if prescription spending in 2006 for a retiree is projected to be $4000, we calculated the mean out-of-pocket spending under a Medicare Part D—like structure as the sum of the $250 deductible, $500 in coinsurance, and $1750 in the doughnut hole, for a total out-of-pocket payment of $2500. Therefore, the generosity of coverage under the Medicare Part D—like structure is computed to be 37.5% (ie, the percentage of drug spending paid by a third party), and the retiree was classified under the 1% to 50% generosity category. We then estimated what essential medication use would be under these new generosity levels. All analyses were performed using STATA software version 8.0 (StataCorp LP, College Station, Tex) and accounted for the complex survey sample design.

Retirees in both disease groups received generous drug benefits from their employer (

). More than three quarters of annual drug expenditures were paid by employer insurance for 54.8% of retirees with CHD and hyperlipidemia and for 44.6% of retirees with CHF. Only 7.1% to 12.2% of retirees in both groups received no retiree coverage for their annual drug expenditures (ie, the 0% group). The differences in the socioeconomic, demographic, and clinical characteristics of the retirees across the generosity of drug coverage categories for the group with CHD and hyperlipidemia and the group with CHF are available in an appendix from the author.

The overall prevalence of statin use was 64.1% among retirees with CHD and hyperlipidemia and that of ACE inhibitor or ARB use was 50.0% among retirees with CHF (Table 1, last 2 columns). Essential cardiovascular medication use was lower for those with less generosity of drug coverage.

For instance, while 72.6% of the most generous drug coverage category (76%-100%) received statins, the prevalence dropped significantly (from 45.6% to 55.5%) in the less generous drug coverage categories. A similar pattern of ACE inhibitor or ARB use by drug coverage generosity was observed among retirees with CHF.

gives the odds ratios for essential cardiovascular medication use by drug coverage generosity and disease group. Given the observed differences in characteristics by drug coverage generosity, logistic regression analyses were used to estimate adjusted odds ratios in addition to unadjusted odds ratios. In general, the magnitude and significance of unadjusted odds ratios remained similar even after adjusting for all covariates. The results confirm that less generosity of drug coverage was associated with lower odds of using an essential cardiovascular medication. In both disease groups, retirees in the less generous drug coverage categories (0%, 1%-50%, and 51%-75%) had significantly lower adjusted odds of essential cardiovascular medication use than retirees with the most generous drug benefits (76%-100%). Also, results of the Durbin-Wu-Hausman test were not statistically significant in either disease group, thus ruling out bias due to unmeasured confounders.

Generosity of coverage was estimated to change substantially in both disease groups if the retiree drug benefits were replaced with standard Medicare Part D—like benefits (the details are available in an appendix from the author). For the retirees in the group with CHD and hyperlipidemia, if those with no drug benefits were to receive Medicare Part D—like benefits, their generosity level would move from 0% to 40%, and if those with the most generous drug coverage were to receive Medicare Part D—like benefits, their mean generosity would move from 85% to 50%. Similar changes were observed in the various generosity of coverage categories among retirees with CHF.

show rates of essential cardiovascular medication use corresponding to the generosity levels under the existing retiree drug benefits and Medicare Part D—like benefits in the 2 disease groups. In the group with CHD and hyperlipidemia, retirees with no drug benefits (0%) in our study period would have a 5.4% increase in statin use under Medicare Part D—like benefits. On the other hand, retirees who had the most generous drug coverage (76%-100%) would have a decrease of almost 16% in statin use under a similar change in structure. Overall, the shift to a standard Medicare Part D—like benefit would result in a mean decline of 8.4% in statin use among all retirees with CHD and hyperlipidemia. We observe similar (albeit smaller) overall effects (-5.2%) among retirees with CHF. However, even in the group with CHF, the most generous drug coverage category (76%-100%) would experience more than double the overall mean decline in ACE inhibitor or ARB use (-11.9%).

Only one half to two thirds of all retirees with the selected heart conditions used essential cardiovascular medications. The rates of use varied substantially based on the generosity of drug benefits among retirees. As expected, retirees with no drug benefits had the lowest rates of essential cardiovascular medication use, followed by those with limited drug benefits. However, even retirees with moderately generous drug benefits had significantly lower odds of using these effective cardiovascular medications than retirees with the most generous drug benefits. This suggests that only very low cost-sharing (ie, 0%-25% of annual drug expenditures paid out of pocket) is associated with a high rate of essential cardiovascular medication use among patients with selected heart conditions.

To our knowledge, this study is among the first to examine the association between generosity of drug coverage and essential medication use specifically among Medicare beneficiaries with retiree health insurance. Our results suggest that retirees who now enjoy generous drug benefits are at risk for reduced utilization of essential medications if their drug coverage generosity was scaled back to the generosity levels seen under Medicare Part D. On the other hand, retirees with previously no drug coverage would have increased utilization of needed medications. Given the recent implementation of Medicare Part D, these findings have important implications for seniors with retiree health insurance who may be at further risk for future reductions in drug coverage generosity.

Our study results must be interpreted in light of the following limitations. First, we classified each retiree into a generosity of drug coverage category based on his or her drug utilization behavior and spending during the year. Because the MCBS data for survey years before 2004 do not have information on the benefit design features of the drug coverage held by each beneficiary, this is the best available proxy for ex ante generosity in the data used for the present study. It is possible that this measure classified people with the same type of coverage into different generosity categories depending on their annual drug use, particularly retirees who faced an annual drug deductible (eg, beneficiaries with extremely generous drug coverage after an initial deductible are misclassified into the 0% generosity category if their annual expenditures were less than the deductible limit). However, according to the Kaiser/Hewitt

2005 survey, only 19% of the employers have a separate annual drug deductible; drug deductibles range from $25 to $250, and the most common deductible is $100. According to established clinical guidelines,

the medications studied herein should be taken chronically, and a 1-year supply of the study medications by themselves would be several hundred dollars. Hence, it is rare that such misclassification would occur in our study given the medication needs of our study samples. Second, because those with greater medication needs may obtain more generous drug coverage, there is a potential for selection bias. However, this is mitigated by the fact that our study sample consisted exclusively of Medicare beneficiaries with retiree health insurance, wherein selection effects are less likely to arise because eligibility is determined more by tenure and industry than by individual self-selection. Moreover, our results suggest that unobserved variables are unlikely to generate biased results because the Durbin-Wu-Hausman test was rejected and because the unadjusted odds ratios changed little even after the numerous controls for observed severity were added to our regression models. Third, our modeling exercise assumed that all retirees would face generosity levels similar to those under Medicare Part D. Although this scenario is unlikely, the objective was to estimate the relative effect of changes in generosity levels on essential medication use, and there is no reason to believe that this should differ markedly from what we have estimated. Fourth, it should be noted that there was only 1 statin (ie, lovastatin) and 1 ACE inhibitor (ie, captopril) and no ARBs available as generics during our study period (1996-2000). Availability of cheaper generics is likely to attenuate our estimated relationship between generosity of drug benefits and use of these cardiovascular medication classes. However, our results are still reflective of the potential effect on use of other newer drug classes without generic substitutes available.

There have been widespread media reports and anecdotal stories of the hardships faced by some retirees with increases in out-of-pocket prescription drug costs, as well as speculations on the potential adverse effects of the new Medicare Part D on employer-sponsored retiree health benefits.

2006 survey of firms with 200 or more employees indicate that about 82% of surveyed employers continued to offer prescription drug coverage while accepting the tax-free subsidy for their largest group of Medicare retirees in 2006. Only 8% of surveyed employers discontinued drug coverage for the plan with the largest group of retirees, and another 2% of surveyed employers have transitioned to a Medicare prescription drug plan. While this is comforting news in the short term, given that most employers today offer drug coverage that is of greater actuarial value than the standard Medicare drug benefit, the trend in shifting costs to retirees is increasing even among employers continuing to offer retiree prescription coverage. For instance, the Kaiser/Hewitt

2006 survey reports that 25% of the employers raised prescription copayments or coinsurance for retirees and that 10% replaced fixed dollar copayments for drugs with coinsurance (ie, retiree pays a percentage of the total drug price) for their retirees, thus exposing them to higher out-of-pocket spending as the costs of drugs rise. Over time, it is possible that additional factors such as accounting requirements, changing demographics, the economy, and labor relations may also influence some employers to considerably scale back drug benefits or to consider other alternatives. How employers will ultimately behave will only become clear during the next few years. Policy makers and researchers must monitor employer response to Medicare Part D during the coming years and determine the extent to which access to effective medications and potentially the health of the nation's retired seniors may be adversely affected.

 From the Division of General Internal Medicine, Department of Medicine, School of Medicine, and the Leonard Davis Institute of Health Economics, University of Pennsylvania, Philadelphia.

 This work was supported by a grant from the Leonard Davis Institute of Health Economics and by American Heart Association Pharmaceutical Roundtable Award 0675066.

 Jalpa A. Doshi, PhD, Division of General Internal Medicine, Department of Medicine, School of Medicine, University of Pennsylvania, 1222 Blockley Hall, Philadelphia, PA 19104-6021. E-mail: jdoshi@mail.med.upenn.edu.

 The authors (JAD, DP) report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter discussed in this manuscript.

 Concept and design (JAD, DP); acquisition of data (JAD); analysis and interpretation of data (JAD, DP); drafting of the manuscript (JAD, DP); critical revision of the manuscript for important intellectual content (JAD, DP); statistical analysis (JAD, DP); provision of study materials or patients (JAD); obtaining funding (JAD, DP); administrative, technical, or logical support; supervision (JAD).

1. McArdle FB, Neuman P, Kitchman M, Kirland K,Yamamoto D. 

Large firmsâ€™ retiree health benefits before Medicare reform: 2003 survey results. 

Employersâ€™ benefits from workersâ€™ health insurance. 

Current trends and future outlook for retiree health benefits: findings from the Kaiser/Hewitt 2004 survey on retiree health benefits. 2004. http://www.kff.org/medicare/7194/index.cfm. Accessed February 15, 2007.

Letter to the Honorable Don Nichols providing additional information about CBOâ€™s cost estimate for the conference agreement on H.R. 1. 2003. http://www.cbo.gov/ftpdocs/ 48xx/doc4814/11-20-MedicareLetter2.pdf. Accessed January 2, 2005

Medicare law is seen leading to cuts in drug benefits for retirees. 

Prospects for retiree health benefits as Medicare pre-scription drug coverage begins: findings from the Kaiser/Hewitt 2005 survey on retiree health benefits. 2005. http://www.kff.org/medicare/ upload/7439.pdf. Accessed February 15, 2007.

Retiree health benefits examined: findings from the Kaiser/Hewitt 2006 survey on retiree health benefits. 2006. http://www. kff.org/Medicare/upload/7587.pdf. Accessed February 16, 2007.

Summary of the Second Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel II). 

Hunt SA, Baker DW, Chin MH, et al; American College of Cardiology/American Heart Association. 

ACC/AHA guidelines for the evaluation and management of chronic heart failure in the adult: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to revise the 1995 Guidelines for the Evaluation and Management of Heart Failure). 

10. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. 

Executive Summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). 

Estimation and Inference in Econometrics. 

New York, NY: Oxford University Press; 1993.

12. Centers for Medicare & Medicaid Services. 

Prescription drug expenditures; aggregate and per capita amounts, percent distribution and average annual percent change by source of funds: selected calendar years 1990-2013. http://www.cms.hhs.gov/NationalHealthExpenData/ downloads/tables.pdf. Accessed February 1, 2006. 

13. Health plans design prescription drug plans tailored to 2007 retiree benefits. 

March 20, 2006. http://www. aishealth.com/ManagedCare/GenBus/MCW_Prescription_Plans.html. Accessed May 31, 2007.

POLICY

Drug Benefit Generosity and Essential Medication Use Among Medicare-eligible Retirees

 To determine whether patient access to secure patientâ€“physician messaging affects annual adult primary care office visit and documented telephone contact rates.

Retrospective cohort and matched-control studies with pre-post analysis.

The cohort study sample included 4686 adult members of Kaiser Permanente Northwest (KPNW) who had been registered KP HealthConnectâ„¢ Online users longer than 13 months and had used at least 1 feature. The matched-control study sample included 3201 randomly selected controls matched by age/sex, selected chronic conditions, and primary care physician to 3201 registered users. We calculated the difference in primary care office visit and documented telephone contact rates in the pre- and post-periods (defined, respectively, as 3-14 months before and 2-13 months after registration for KP HealthConnectâ„¢ Online). Paired t tests were used to assess significance.

 Annual office visit rates decreased by 0.23 (-9.7%) visits per member in the cohort study. Annual office visit rates for users in the matched-control study decreased by 0.25 (-10.3%); the corresponding decrease for the controls was 0.08 (-3.7%). This 0.17 (-6.7%) reduction was significant (P < .003). Annual documented telephone contact rates for users in the matched-control design increased by 0.32 (16.2%) contacts per member; the corresponding rate for the control group was 0.52 (29.9%). This 0.20 (13.7%) difference was significant (P < .01).

Patient access to the secure messaging feature of KP HealthConnectâ„¢ Online was associated with decreased rates of primary care office visits and telephone contacts.

  This is the largest study to date of the impact of access to secure patientâ€“physician messaging on provider workload. The results demonstrated that:  

Annual adult primary care outpatient visits decreased by 6.7% to 9.7% among patients using electronic messaging.

Members using electronic messaging had a smaller increase in documented telephone contacts (16.2%) than the control group (29.9%).

Electronic messaging may provide a solution to pervasive efficiency and access issues for both patients and providers.

Widespread efforts to improve healthcare quality, safety, and efficiency focus on using information technologies such as electronic health records, patient registries, computerized physician order entry, embedded decision supports, and others. Among these, electronic patient—physician messaging has been viewed as a promising technology to improve the quality and efficiency of healthcare.

Electronic communication between patients and physicians can reduce health plan spending on physician office and laboratory services.

Patients and physicians alike indicate satisfaction with electronic messaging.

Ninety percent of US consumers with Internet access indicate a clear preference for online communications with healthcare providers.

Complex issues of reimbursement, confidentiality, and liability have impeded widespread adoption of patient—physician electronic messaging in the United States. However, these issues appear to be resolving. Following the American Medical Association's 2004 approval of online consultation billing, some US insurers have recently offered reimbursement for its use.

Similarly, secure Web messaging represents a security improvement over e-mail, reduces liability,

and is increasingly available as a stand-alone service to physicians without electronic health record systems.

Some physicians may be concerned that the use of secure messaging would increase their overall workload. 

The extent to which secure messaging can substitute for office visits or telephone contacts is unclear. Previous studies offer conflicting evidence: telephone communication is not impacted by electronic messaging,

and both telephone contacts and office visits are reduced by electronic communication.

These studies were small in size and limited in scope; the largest involved roughly 1000 health-related messages.

To investigate the relationship between patient—physician electronic messaging and physician workload, we evaluated the impact of patient access to an electronic health record with secure patient—physician messaging on primary care office visit and documented telephone contact rates in an entire Kaiser Permanente (KP) operating region.

KP is the nation's largest not-for-profit integrated healthcare delivery system with more than 8.5 million members in 8 geographic regions. Professional partnerships in each region employ physicians and contract with the not-for-profit Kaiser Foundation Health Plan to arrange necessary medical care for members.

KP's integrated healthcare delivery system addresses all healthcare needs for adult and pediatric members, including preventive, routine, specialty, emergency, and inpatient care; ancillary testing; pharmacy and rehabilitative services; and home care. The Kaiser Permanente Northwest (KPNW) region, with nearly 487 000 adult and pediatric members in April 2006, is located in Oregon and southwest Washington. 

 KP is implementing an integrated electronic health record throughout the entire enterprise.

Based on software supplied by Epic Systems, it is known as KP HealthConnectâ„¢. Members can access parts of their individual health records through a secure member Web site: www.kp.org. After registering as users, they may take advantage of all KP HealthConnectâ„¢ Online features (

). Messaging takes place within a secure Web environment. User accounts are activated by passwords mailed to members' homes, and all messaging takes place in an authenticated/encrypted environment and behind an enterprise-level firewall.

Members are clearly informed that receiving a response to a secure message may take up to 2 business days and that messaging is only appropriate for nonurgent concerns. The member-only Web site, www.kp.org, contains the following warnings on the secure messaging screen: "Do not attempt to access emergency care through this Web site … call 911 or go to the nearest hospital." "If you have an urgent symptom or want to speak with a nurse, do not use this Web site. Please call your local Kaiser Permanente facility." 

KP HealthConnectâ„¢ Online was first implemented as a pilot project (Personal Health Link) in November 2002 for adult members in the Northwest region. As of September 2005, 18 094 members were registered users, representing 6% of the total KPNW adult membership. 

 To evaluate the impact of KP HealthConnectâ„¢ Online on primary care office visit and documented telephone contact rates, we conducted a pair of retrospective studies. The 2 designs relied on administrative data about subject characteristics, primary care office visit and telephone contact rates, and KP HealthConnectâ„¢ Online use. The use of complementary designs allowed us to examine the impact of secure messaging in 2 ways: in the largest possible sample and while controlling for factors that might impact validity.

A retrospective cohort study included 4686 adult members who were registered KP HealthConnectâ„¢ Online users for longer than 13 months, had used at least 1 feature, and were continuously enrolled as KPNW members during the study period. Cohort subjects must have registered to use KP HealthConnectâ„¢ Online by August 2004. For the majority, the study period occurred between September 2002 and August 2005. Cohort subjects were on the patient panels of approximately 250 primary care physicians.

A retrospective matched-control study included 3201 subjects who were also part of the cohort described above. For each, we identified and randomly selected control matched by age, sex, selected chronic conditions (eg, diabetes, congestive heart failure), and primary care physician.

 For both studies, we defined the pre-period as 3 to 14 months before KP HealthConnectâ„¢ Online registration and the post-period as 2 to 13 months afterward. Immediately around the time of registration, outpatient visit rates for subjects in both the cohort and matched-control studies were above baseline levels because many people learned about KP HealthConnectâ„¢ Online during clinic visits for active health concerns. To conservatively estimate the impact on utilization of access to KP HealthConnectâ„¢ Online, we omitted this spike in outpatient visit and telephone contact rates from our analysis.

 Outcome variables included annual adult primary care office visit rates, comprising appointments with physicians and physician extenders (nurse practitioners and physician assistants) in adult primary and urgent care (nonemergency) settings. Documented telephone contact rates included both scheduled telephone visits and unscheduled telephone calls to and from internal medicine and family practice physicians, nurse practitioners, and physician assistants. 

test to look for differences in age/sex and the prevalence of diabetes and congestive heart failure between 5 groups of interest: the 323 296 adult members of KPNW, 18 094 KPNW members who had registered with KP HealthConnectâ„¢ as of September 2005, the 4686 cohort subjects, the 3201 subjects of the matched-control study, and the 3201 controls of the matched-control study. Differences in office visit and telephone call rates were symmetric about the mean, although not normally distributed based on the formal normality test. However, because percent changes in mean utilization rates are the most operationally meaningful way of examining trends, and parametric tests are robust to deviations from Gaussian distributions when samples are large,

test to assess the statistical significance of differences in utilization rates over time and across groups. As a matter of interest, the Wilcoxon rank sum test and the paired 

test yielded identical statistical significance. 

We calculated the difference in primary care office visit and documented telephone contact rates between the pre- and post-periods, assessing statistical significance with the paired 

We calculated the difference in primary care office visit and documented telephone contact rates in the pre- and post-periods for subjects and for controls, again assessing statistical significance with the paired 

test. In addition, we also used the paired 

test to assess the statistical significance of the variation in rate changes between the subject and control groups. 

 In general, KP HealthConnectâ„¢ Online users were older and included a higher proportion of members with diabetes than did the general adult membership. Similarly, cohort members were older and included a higher proportion of individuals with diabetes than did the larger user population. The differences in age/sex and the proportion of members with diabetes between the cohort and the general adult membership were significant (

< .0001). By design, cohort subjects and both groups in the matched-control study did not differ to a statistically significant degree in terms of age and the proportion of members with diabetes. 

Annual Adult Primary Care Office Visit Rates

 Baseline primary care office visit rates for cohort and matched-control subjects were slightly higher than the baseline rates for the entire region, consistent with the increased prevalence of chronic conditions. The baseline visit rate of the control group in the matched-control study was between the regional baseline rate and the subjects' rate. 

Annual adult primary care office visit rates decreased by 9.7%, a statistically significant decline from 2.47 to 2.24 office visits per member per year (

For the subjects in the matched-control study, the annual adult primary care office visit rate decreased by 10.3%, or 0.25 visits per member per year (

< .001). The corresponding decrease for controls was 3.7%, or 0.08 visits (

< .003). The difference between changes in primary care office visit rates for the 2 groups, 6.7%, was also statistically significant (

summarizes office visit utilization results for both studies.

 KPNW implemented new documentation procedures for telephone contacts during the study period; as a result, documented primary care telephone contact rates for the entire region increased by 24%. Documented primary care telephone rates for the cohort subjects increased by 15.6%. To evaluate the impact of access to KP HealthConnectâ„¢ Online in the context of this broad trend, we relied on the matched-control study. 

The annual primary care telephone contact rate for subjects increased by 16.2%, or 0.32 documented telephone contacts per member per year, over the study period (

< .001). The corresponding increase for controls was 29.9%, or 0.52 documented telephone contacts per member per year (

< .001). The difference between these increases (13.7%) was also statistically significant (

 We evaluated the impact on office visit and telephone contact rates of patient access to an integrated multifunction electronic personal health record that included secure patient—physician electronic messaging. Annual adult primary care outpatient visit rates decreased by 6.7% to 9.7% for members using KP HealthConnectâ„¢ Online, and these members had a smaller increase in documented telephone contacts (16.2%) than the control group (29.9%). 

 Conducting a randomized controlled trial would have required fundamental changes to the KP HealthConnectâ„¢ system so that only patients randomized to the intervention were allowed to access online features. However, inconsistent member access to system features would have resulted in a prohibitive work-flow burden for physicians and healthcare teams. Additionally, random implementation of secure messaging would have confounded its true operational impact on office visit utilization. Nevertheless, our study controlled for individual patient factors, physician work styles, and regional trends as alternative explanations for reduced utilization.

Access to parts of the personal health record or other KP HealthConnectâ„¢ Online features may have influenced primary care office visit and documented telephone contact rates. However, an early evaluation of KP HealthConnectâ„¢

Online use among 1000 registered users found that more than 70% of sessions resulted in patient—physician messaging, indicating the importance and influence of this feature. 

Although our sample size didn't support evaluating the impact of individual features, registered users most frequently cited telephone calls and office visits as alternatives to secure messaging. A random sample of 2700 KP HealthConnectâ„¢ Online users who e-mailed their physicians during a 3-month period yielded more than 1700 completed questionnaires. A quarter of the respondents indicated they would have scheduled an appointment in lieu of electronic messaging (

) and were satisfied with and appreciated the alternative mode of care.

A limitation of our study is that the subjects and controls in the matched-control study were not paired by baseline office visit or telephone contact rates. Subjects had higher preregistration utilization rates than did controls because high utilizers were more likely to register. Further research would assess whether different baseline utilization rates affect the impact of access to secure messaging on utilization rates. 

Our study suggests several additional areas for further study. Annual primary care office visit rates held steady for the region as a whole. However, visit rates were lower, to a statistically significant degree, in the post-period for both groups in the matched-control study. We hypothesize that, because subjects and controls were matched by primary care physician (among other characteristics), these physicians may have become more attuned to care efficiencies during the study period. Further research would validate this hypothesis. 

Members with diabetes were disproportionately represented among KP HealthConnectâ„¢ Online users. This fact raises important questions about electronic communications in chronic illness care. Other models of electronic communications, such as Internet-based glucose self-monitoring programs, have proved effective in increasing glucose control over the short term,

and physicians view electronic communication as enhancing care for patients with chronic conditions.

Anecdotal evidence from KP physicians indicates that secure messaging may actually increase the number of patient "touches." In addition to messaging itself, in 19.4% of 93 randomly sampled patient—physician secure message threads, the physician recommended an additional contact: a laboratory test, phone call, office visit, procedure, or health education class. It is important to note that some key elements of diabetes clinical treatment guidelines (ie, glycated hemoglobin [HbA1C], lipid, and microalbuminuria monitoring) do not require a face-to-face physician office visit with the availability of an electronic medical record. 

To confirm that secure messaging is used for nonurgent issues, a review of the level of service of 50 secure messaging threads revealed that two thirds were coded as either "brief" or lower.

KPNW collects data for the Health Employer Data and Information Set (HEDIS

) as part of routine quality surveillance. The HEDIS reports for HbA1c testing did not vary to a statistically significant degree during the years under observation. 

Secure messaging reduces overall physician workload if it requires less time than the replaced visits and telephone contacts. Although we did not examine overall efficiency, patients perceive electronic messaging as preferable to telephone consultations in many situations.

Physicians and staff state that electronic messaging requires less time than telephone calls and that lengthy messages can be completed at intervals throughout the day.

The extra capacity that secure messaging creates through increased efficiency can be used at the discretion of the care provider or organization. In noncapitated systems, an overall reduction in office visit rates may not be financially advantageous, and providers may choose to fill the resulting extra capacity with additional patient visits with a higher level of service to recover lost revenue. In a system with different incentives, more preventive care could take place at each visit or the time freed up by reduced office visit rates could be used for panel management.

KP is a largely prepaid, integrated healthcare delivery system. Patients and physicians generally used electronic messaging free of the reimbursement concerns that presently challenge the US healthcare system. The results from this large-scale study of secure messaging indicate that, as these issues resolve, it may provide a win-win-win solution to pervasive efficiency and access issues from the perspectives of patient, healthcare provider, and payer.

The authors would like to acknowledge many physicians, operation lead ers, and analysts in the Kaiser Permanente Northwest region and Program Offices for their work. In particular, we thank Susan Moy, MS, Carl Serrato, PhD, Sharon Fox, MS, Rene Walton, MS, Michael Kositch, MD, Robert Peterson, MD, David Shenson, MD, Anna-Lise Silvestre, MPH, and Kate Christensen, MD, for their leadership and many contributions, and Jenni Green, MS, for advice and help in writing this article. 

Kaiser Foundation Health Planâ€”program offices (YYZ, TG, LLL); Northwest Permanente (HLC); The Permanente Federation (AMW). 

Yi Yvonne Zhou, PhD, Kaiser Permanente, 2850 NW Nicolai St, Portland, OR 97210. E-mail: yvonne.y.zhou@kp.org.

The authors (YYZ, TG, HLC, AMW, LLL) report relationship or financial interest with any entity that would pose a conflict interest with the subject matter discussed in this manuscript.

Concept and design (YYZ, TG, HLC, AMW, LLL); acquisition of data (YYZ); analysis and interpretation of data (YYZ, TG, HLC); drafting of the manuscript (YYZ, TG); and critical revision of the man uscript for important intellectual content (YYZ, TG, HLC, AMW, LLL); statis tical analysis (YYZ); obtaining funding (TG, LLL); administrative, technical, logical support (TG, HLC); and supervision (YYZ, TG, HLC, AMW).

Crossing the Quality Chasm: A New Health System for the 21st Century. 

MANAGERIAL

Patient Access to an Electronic Health Record With Secure Messaging: Impact on Primary Care Utilization

 To study the effect of cost-sharing policy changes on utilization of test strips for self-monitoring of blood glucose.

 A legislative mandate (January 1, 2000) required California health plans to cover diabetes supplies, including those for self-monitoring of blood glucose. One health plan, Kaiser Permanente Northern California, initially waived established copayments and provided free test strips to members with diabetes mellitus for 2 years but later instituted a 20% coinsurance charge for a portion of their membership. 

 A retrospective cohort design was used to study pharmacy-based estimates of test strip utilization changes during this natural experiment. Analyses included 2 cohort investigations using pretest-posttest analysis with control subjects to study transitions from a copayment period to a free test strip period and from the free test strip period to a coinsurance period.

 During the copayment period, test strip utilization was inversely related to copayments for test strips. Offering free test strips did not increase utilization, even among those paying higher copayments before the policy change. Price-elastic patterns formed before and during the copayment period persisted, despite receiving free test strips for 2 years. The coinsurance, introduced after 2 years of receiving free test strips, resulted in statistically significant (but not clinically relevant) decreased utilization (approximately 1-3 fewer test strips/month). Change patterns did not differ by socioeconomic status.

 Offering free test strips shifted costs from patient to health plan, without improving adherence. The introduced coinsurance slightly reduced utilization and adherence to recommendations about self-monitoring of blood glucose. Neither intervention had marked clinical effect. Cross-sectional analyses should not be used to predict utilization changes in the face of rapidly evolving benefit policies.

We evaluated changes in utilization of test strips for self-monitoring of blood glucose levels across the following 2 cost-sharing thresholds: (1) from a period of copayments to a period of free strips and then (2) from the period of free strips to a period of coinsurance.

Findings included slightly reduced utilization in response to increased cost sharing but unchanged utilization in response to free test strips.

Price-elastic patterns formed before and during the period of copay-ments persisted, despite receiving free test strips for 2 years.

Cross-sectional analyses should not be used to predict utilization changes in response to changes in benefit policies.

The Diabetes Control and Complications Trial

and the United Kingdom Prospective Diabetes Study

demonstrated the benefits of achieving tight glycemic control in patients with diabetes mellitus (DM) and led to a concomitant promotion of self-monitoring of blood glucose (SMBG) as an integral part of good diabetes care. However, SMBG is consistently practiced less than is recommended by commonly supported guidelines,

including those of the American Diabetes Association.

Test strips for SMBG are expensive, and SMBG practice is more sensitive to cost sharing

(ie, price-elastic demand) than many other health services (eg, dilated eye examinations).

However, current understanding regarding how financial barriers affect utilization is primarily based on cross-sectional analyses.

California Senate Bill 64 (SB64), implemented on January 1, 2000, mandated that California's health plans provide coverage for diabetes supplies, including those for SMBG. The bill was intended to eliminate a financial barrier to SMBG and to facilitate this self-care practice. In recent years, 37 other states have passed similar legislation. Following similar legislation, rates of SMBG utilization increased, and glycemic control improved after a New England health plan started providing free glucose monitors in 1993.

In the year before implementation of SB64 (hereafter referred to as the copayment period), a California health plan, Kaiser Permanente Northern California (KPNC), charged members with DM a $0 to $30 copayment for each prescription of test strips, depending on the member's type of benefit coverage. During the 2 years after SB64 was implemented (January 1, 2000, through December 31, 2001) (hereafter referred to as the free test strip period), test strips were provided free of charge to all KPNC members with DM. A second policy change occurred on January 1, 2002, after which approximately 35% of the commercial membership were charged a 20% coinsurance (approximately $8-$12 per prescription, depending on the brand of test strips) (hereafter referred to as the coinsurance period).

We studied the effects of these 2 benefit policy changes on SMBG practice (as measured by test strip utilization). We hypothesized an increase in individual-level test strip utilization after test strips were provided for free, with the greatest increase among those paying the largest copayments before the implementation of SB64. We assumed that poorer members would be more price sensitive and hypothesized that provision of free test strips would reduce socioeconomic disparities in utilization. We also hypothesized that, after receiving free test strips for 2 years, utilization would decrease after the coinsurance was introduced and that this decrease would be greatest among the poor and among those not treated with insulin.

Kaiser Permanente Northern California is a group-model healthcare organization that provides comprehensive medical services to more than 3 million members (approximately 35% of the population in the surrounding geographic area) through 15 hospitals and 23 outpatient clinics. Study subjects were identified from the KPNC diabetes registry (approximately 132 000 in 2000).

Two cohorts were created to analyze the effect of the 2 separate policy changes. The first was created for the analysis of utilization before and after the introduction of free test strips (hereafter referred to as the free test strip cohort), and the second was created to assess utilization before and after institution of a coinsurance for test strips (hereafter referred to as the coinsurance cohort). For each cohort, we created a sampling frame that included subjects with continuous health plan membership and prescription drug benefits in the prepolicy and postpolicy periods who were also ongoing test strip users (defined as having at least 1 dispensing for test strips during each year of the study). We did not include members without prescription drug benefits and those with prescription drug caps, because they have little financial incentive to purchase SMBG supplies exclusively at KPNC pharmacies and we cannot ascertain utilization at non-KPNC pharmacies. Based on these inclusion criteria, we identified 33 596 subjects in the free test strip cohort and 27 810 subjects in the coinsurance cohort sampling frames. We then excluded 12 241 and 9841 noncommercial health plan members (ie, members in subsidized plans [eg, Medicare]) from the free test strip cohort and the coinsurance cohort, respectively, to reduce case mix, demographic, and socioeconomic heterogeneity in the cohort. Patients enrolled in Medicare received free test strips starting July 1, 1998; this age class would have dominated the reference group if included.

Pharmacy medication dispensing records were then used to classify subjects into the following 3 mutually exclusive treatment groups: patients who use any regimen that includes insulin (hereafter referred to as the insulin group), patients treated only with oral hypoglycemic agents (OHAs) (hereafter referred to as the OHA group), and patients receiving no diabetes pharmacotherapy but receiving medical nutrition therapy (MNT) only (hereafter referred to as the MNT group). Healthcare providers generally encourage increases in SMBG frequency with the intensification of or the introduction of new pharmacotherapy; in these individuals, we would expect an increase in SMBG frequency during this time, independent of the change in the cost of test strips. Therefore, we excluded 2107 and 2323 additional subjects from the free test strip cohort and the coinsurance cohort, respectively, who switched type of diabetes therapy modalities (eg, adding insulin to regimens that previously included oral agents only) during the study period, although we could not exclude those with dosing changes or with additions of other therapies within the oral agent or insulin modality. The final free test strip cohort and coinsurance cohort on which this study was based included 19 248 and 15 646 subjects, respectively.

Because the 2 benefit changes in charges for test strips were not announced in writing to the KPNC membership, only ongoing users were aware of price changes (via the change in out-of-pocket charge for test strips). Therefore, for our main analyses of the individual-level effect of changes in out-of-pocket cost (eg, pre-post policy), we included only test strip users at baseline whom we assumed had experienced the changes in charges for test strips. However, although exclusion of individuals who were not prevalent users was appropriate for assessing individual-level changes, these cohorts are not well suited to study population-level patterns of utilization across levels of out-of-pocket cost because high cost may also lead to never initiating use or to discontinuing use. Therefore, we evaluated trends in SMBG practice and price sensitivity in a panel study of annual cohorts that included the individuals who were nonusers.

We assessed SMBG practice as indicated by test strip utilization (mean daily test strip utilization) recorded from pharmacy dispensing records. The use of pharmacy dispensing records to estimate utilization has been shown to be a valid measure for patients with a prescription benefit.

For each calendar year, the mean number of test strips used per day was calculated for each member by summing the cumulative number of test strips for the year and dividing by 365 days. We assumed that a patient possessed no stockpile of test strips at baseline. Test strip copayment amounts paid out-of-pocket were available through KPNC's benefits database. Copayment levels were collapsed into the following 4 categories: $0 (reference group), $1 to $3, $4 to $9, and $10 and higher to identify price-elastic patterns. Neighborhood-level socioeconomic status was characterized by geocoding each member's address to its 2000 census block group and by creating variables for median household income in 1999 and for residence in an impoverished neighborhood (defined as >20% of the population with an income below the US poverty line).

We first detailed population-level 1995 to 2002 secular trends in SMBG (unadjusted), defined as a continuous measure (mean number of test strips used per day) and as a discrete measure (prevalent use vs nonuse). We then used a longitudinal retrospective cohort investigation in 2 separate cohorts to evaluate longitudinal changes in test strip utilization spanning a 4-year period across the following 2 thresholds: (1) transition from the copayment period to the free test strip period and then (2) transition from the free test strip to the coinsur-ance period. The outcome of interest was the change in the mean number of test strips used per day associated with the policy-based health plan—wide change in cost sharing. To control for selection bias and for intervening time factors (secular trends and aging), we used the difference-in-differences method (pretest-posttest analysis with control subjects).

Hierarchical linear models (random coefficient models) were specified to determine the 1-year change in the mean daily test strip utilization following the introduction of 2 separate cost-sharing policies. All statistical analyses were conducted using SAS software (SAS Institute, Cary, NC); the multivariate models were run using PROC MIXED (SAS Institute). The reference group, the $0 copayment group, received free test strips before and after each policy change and was used to account for the secular trend in SMBG in the absence of any price change. We expected that the levels of SMBG and the patterns of change in test strip utilization might cluster within the different medical facilities, which may have differed in their efforts to promote SMBG; therefore, we specified a random intercept for medical facility in our statistical models.

We assumed a priori whose effects would be modified by diabetes therapy, so we specified separate models for the insulin, OHA, and MNT groups. The exposures of interest (benefit-based changes in test strip copayment) were exogenous policy changes and were unlikely to be related to any individual-level factors. To be sure, we also specified adjusted models that controlled for potentially confounding individual-level prebaseline variables. These covariates included the mean daily test strip utilization during the year before the policy change, prepolicy prescription drug copayment (for the analysis of the first policy change only), prepolicy glycosylat-ed hemoglobin level, age, missed appointment rate, sex, inpa-tient comorbidity score, total pharmacy cost in the prebaseline year, specialty visits, urgent care visits, primary care visits, residence in a predominantly working-class occupation neighborhood, residence in a predominantly undered-ucated neighborhood, block group median income (1999), oral medication refill adherence (OHA models only), and pharmacy-derived utilization of insulin syringes as a measure of daily insulin injection frequency (insulin models only). We found no substantive difference between unadjusted and adjusted models; therefore, we present only the unadjusted findings. In addition, we specified models with cross-product terms stratified by contextual measures of socioeconomic status, as we hypothesized that residence in an impoverished area might modify the degree of changes in SMBG after the implementation of each policy.

Price Elasticity and Population Secular Trends

We evaluated the population-level secular trends in SMBG as a discrete measure (prevalent use vs nonuse) (

) and as a continuous measure (mean number of test strips used per day) (data not shown) between 1995 and 2002 using the cohorts that did not exclude nonusers or inconsistent test strip users. There was a gradual trend toward higher prevalence of SMBG and more test strips used per day during the observation window. Consistent with price-elastic utilization, the mean daily test strip use and the prevalence of SMBG were consistently lower among those paying the higher out-of-pocket copayment at any point in time. Moreover, although the copayment charges that applied during the copayment period (1995-1999) were no longer in effect after 1999, the price-elastic patterns formed during the copayment period persisted.

Baseline Characteristics of the Free Test Strip Cohort

We identified 19 248 members with DM in the free test strip cohort who had pharmacy benefits, were test strip users during the year preceding the implementation of SB64, and did not change diabetes therapy type during the follow-up period. The reference group for the free test strip cohort comprised approximately 7% of the subjects who received test strips for free before the policy change (

). The remaining subjects were distributed across different levels of copay-ment. The free test strip cohort included the following 3 treatment strata: 8151 subjects in the insulin group, 10 049 subjects in the OHA group, and 1048 subjects in the MNT group. Given the large sample sizes, it is important to focus attention on clinically relevant differences, despite the several statistically significant differences across copayment groups. Diabetes therapy groups and the mean use of SMBG were similarly distributed across copayment groups.

Test Strip Utilization Following the Introduction of Free Test Strips

Parameter estimates for the 1-year change in the mean daily test strip utilization following the introduction of free test strips, after adjusting for secular trends observed in the reference group, were inconsistent with our study hypothesis (

). We had hypothesized that SMBG would increase after the provision of free test strips following the policy change and that the degree of this increase would be proportional to the amount of out-of-pocket cost paid per test strip prescription during the year before this policy change. In the OHA and MNT groups, there were no substantive or statistically significant increases in SMBG relative to the control group. There was no evidence of price elasticity in response to the change in cost in either group. In the insulin group, providing free test strips was accompanied by a counterintuitive decrease, albeit small (range, <0.5-2.5 fewer test strips per month), in SMBG among all copayment groups relative to the reference group.

Although the change in cost sharing (ie, each member's preexisting copayment amount) should not be affected by individual-level variables, we ran adjusted models to rule out potential confounding. The parameter estimates did not change substantively with adjustment, so we present the results from the more parsimonious (unadjusted) models. Furthermore, there was no evidence that socioeconomic status modified our findings, as there were no statistically significant differences in price elasticity among those who lived in predominantly impoverished vs nonimpoverished neighborhoods.

Test Strip Utilization Following the Introduction of a 20% Coinsurance

There were 15 646 subjects in the coinsurance cohort: 4505 subjects in the insulin group, 9627 subjects in the OHA group, and 1514 subjects in the MNT group. Fifty-eight percent of the cohort continued to receive test strips for free after the policy change, and 42% were charged the coinsur-ance ($8-$12). Subject characteristics did not differ clinically among those who were exposed vs those who were unexposed to the second policy change (a 20% coinsurance per prescription), although many of the contrasts were statistically significant because of the large sample sizes (

As in the previous analysis, we ran simple and adjusted models to determine the effect of this policy change on SMBG. Adjustment for individual-level covariates did not alter our results (ie, they were not confounders), so we report the results of models that adjust for prepolicy SMBG only. In the year following the policy change, subjects in the insulin, OHA, and MNT groups who were charged the coinsurance decreased their SMBG on average by 0.09 (

= .04) test strips per day, respectively, compared with those who continued to receive test strips for free. Although statistically significant, this decrease (approximately 2 test strips/month) is too small to be of clinical relevance.

In addition, there was no statistically significant difference in the SMBG discontinuation rates between those who were and those who were not charged the coinsurance. There was no statistically significant difference in the degree of the decline by socioeconomic status (ie, those living in vs those not living in impoverished neighborhoods) (data not shown).

This natural experiment provided a unique opportunity to evaluate the effect of changes in out-of-pocket costs on individual and population-level utilization of SMBG. We observed the effect of 2 policy-driven changes in cost sharing on SMBG (measured as test strip utilization) in a large managed care population with DM. As expected, we observed price-elastic demand in cross-sectional assessment of SMBG and out-of-pocket costs among prevalent SMBG users, suggesting the presence of a financial barrier to SMBG utilization. This graded utilization pattern was less marked when we restricted our analysis to the subset of patients with DM who were continuous SMBG users. This suggests that there may be a price threshold over which individuals choose to discontinue SMBG practice altogether, rather than to incrementally reduce test strip utilization (and SMBG frequency) in response to increasing test strip costs. Moreover, once a patient decides to initiate SMBG practice, he or she may be less price sensitive.

Our study findings are often counterintuitive and provide new insight into the effect of changes in healthcare policy, cost sharing, and provision of free services on this diabetes self-management practice. We expected that the elimination of the copayment would stimulate greater utilization, especially among those charged the most in the prepolicy period or among the poor. Paradoxically, the offering of free test strips following the implementation of SB64 did not increase SMBG above and beyond that expected by secular changes, despite historical evidence of price-elastic demand (based on the cross-sectional patterns). In fact, insulin users decreased test strip use slightly. Even economically disadvantaged patients and those paying the most for test strips before the policy change did not take extra advantage of the free test strip offer.

The SMBG practice patterns formed during the copay-ment period (before the implementation of SB64) persisted, despite receiving free test strips for 2 years. These entrenched patterns suggest that monitoring frequency may be established at the initiation of SMBG practice, affected by cost and other previously demonstrated predictors such as socioeconomic status, inconvenience, physical discomfort associated with the fingersticks, disease severity, perceived health benefits, provider recommendations, and family support.

However, once individuals have habituated to a stable SMBG practice, it is likely to be less quickly altered by access to free test strips or by even slight increases in cost. It is important to note that the health economic concern that providing free test strips would lead to a spike in excessive utilization (eg, multiple daily tests for oral agent only—treated patients) was unfounded; the prevalence of excessive use of SMBG continued to be dwarfed by the prevalence of underuse (<5% overutilized vs 40% did not practice at all in 2002 based on American Diabetes Association

clinical recommendations) after providing free test strips. Although the introduction of a small 20% coinsurance ($8-$12), the second policy change in the same population, resulted in a statistically significant decline in SMBG, it was too slight to be considered clinically relevant, and the decline did not differ by socioeconomic status.

July 2007

Szpunar et al (Am J Manag Care. 2007;13[pt1]:313-315)

Liu et al (Am J Manag Care. 2007;13[suppl 3]:S80)

Drug Benefit Generosity and Essential Medication Use Among Medicare-eligible Retirees

Patient Access to an Electronic Health Record With Secure Messaging: Impact on Primary Care Utilization

Effect of Cost-sharing Changes on Self-monitoring of Blood Glucose

Quality of Clinical and Economic Evidence in Dossier Formulary Submissions

Patient Preferences for Colorectal Cancer Screening: How Does Stool DNA Testing Fare?

Patient Preferences for Colon Cancer Screening: The Role of Out-of-Pocket Costs