Mapping Geographic Areas of High and Low Drug Adherence in Patients Prescribed Continuing Treatment for Acute Coronary Syndrome After Discharge

Cuong Hoang, Pharm.D.; Giselle Kolenic, M.S.; Eva Kline-Rogers, R.N., N.P.; Kim A. Eagle, M.D.; Steven R. Erickson, Pharm.D.


Pharmacotherapy. 2011;31(10):927-933. 

In This Article

Abstract and Introduction


Study Objective. To determine the feasibility of using geographic information system (GIS) technology to identify geographic areas of high and low adherence to cardiovascular drug therapy for treatment of acute coronary syndrome (ACS) in patients discharged from a university-affiliated hospital.
Design. Retrospective analysis.
Data Source. A registry of patients admitted to and discharged from a large university-affiliated medical center for the treatment of ACS.
Patients. A total of 1081 adults distributed over 300 census tracts who were discharged between April 1999 and December 2004 with a diagnosis of an ACS event of unstable angina or acute myocardial infarction.
Measurements and Main Results. Data were collected on patient demographics, home addresses, and adherence to four classes of drugs—statins, angiotensin-converting enzyme inhibitors, _-blockers, and aspirin—at 6–12 months after discharge for the ACS index event. A GIS program was used to map patient addresses and adherence data to geographic coordinates. Hot Spot Analysis was used to determine the existence of any spatial clustering patterns in adherence rates. The analysis was performed at the census tract level by using the percentage of nonadherent patients within a census tract to represent adherence for the people living within that tract, standardized by the number of residents in a census tract aged 40 years or older. Hot Spot Analysis identified unique geographic areas of high, neutral, and low adherence in the southeast area. Highly adherent census tracts were primarily located in and around the city where the university hospital and clinics are located. Areas of low adherence were located to the west, southwest, and southeast of the city. All other census tracts were considered neutral in adherence rates.
Conclusion. Mapping geographic areas of drug adherence is feasible with use of GIS technology, with spatial mapping able to detect areas of varying levels of adherence. Future research should examine local-level factors associated with low adherence, which can be used to derive tailored, locally relevant interventions to improve long-term drug adherence.


Acute coronary syndrome (ACS) encompasses several clinical conditions, including ST-segment elevation myocardial infarction (STEMI), non–ST-segment elevation myocardial infarction (NSTEMI), and unstable angina. In 2004, the American College of Cardiology and the American Heart Association published evidencebased guidelines for the treatment of ACS.[1] Based on evidence from randomized controlled trials, four classes of drugs are recommended for the treatment of patients with ACS: angiotension-converting enzyme (ACE) inhibitors, β-blockers, statins, and antiplatelet drugs. Since the publication of these guidelines, changes in pharmacologic therapy and interventions for ACS have significantly reduced the number of inhospital deaths, cardiogenic shock, recurrent myocardial infarction, and heart failure in patients with STEMI and NSTEMI.[2]

On discharge, these drugs should be continued to prevent secondary cardiovascular events. However, nonadherence is very common in patients with cardiovascular disease, often leading to increased mortality and hospitalizations.[3–5] Nonadherence has been shown to significantly increase the likelihood of death within 1 year after a myocardial infarction.[6] For example, higher rates of death and acute myocardial infarction were observed in patients who prematurely stopped treatment with clopidogrel.[7]

There are many patient-specific factors that contribute to nonadherence, some of which include age and race-ethnicity, polypharmacy, frequency of drug changes, socioeconomic status, access to medical care, out-of-pocket drug cost, lack of prescription drug insurance, and poor communication between patients and health care providers.[8–13] In addition, beliefs and attitudes about illness and drugs influence adherence to drug regimens.[8,14] Patients may alter their drug regimens based on their perceptions of the effectiveness of the drugs, the constraints of everyday life, and any adverse effects they experience.[15,16]

The focus of research on medication-taking behavior has been on patient-level factors. Identification of these important variables often leads to development of patient-centered interventions tailored to the patient's specific needs. An approach to modifying behavior used in the public health arena is to target populations and their environments rather than individuals.[17] Researchers are beginning to examine the environment in which individuals live in an attempt to identify determinants of illness, as well as the existence of modifiable risk behaviors in populations. Individuals of low socioeconomic position are at greater risk of developing cardiovascular disease, and those living in deprived neighborhoods have lower survival rates after an acute myocardial infarction.[18,19]

Individual factors associated with nonadherence may also be linked to an individual's neighborhood or community. Social characteristics of residential environments are associated with the conduct of healthy behaviors.[20] Evidence supports the concept that residential neighborhoods play a role in determining individual behaviors linked to health outcomes, which primarily have been studied in diet and exercise.[21,22] It may be hypothesized that medication-taking behavior is also linked to neighborhood or geographic characteristics. Recent studies have documented that drug prescribing and adherence vary based on geography. For example, adherence to drugs used for the treatment of diabetes mellitus varies by region in which the patient lives in the United States.[23] Other studies have documented geographic variation in the prescribing of drugs such as opiates and antiretroviral treatment.[24,25]

A relatively new analytic technique, spatial epidemiology, combines the disciplines of geography, statistics, and epidemiology.[26] Among the tools used in these analyses are mapping computer programs. Programs known as geographic information systems (GIS) are capable of not only mapping any type of data that can be linked to an address, or geocoded, but also performing statistical analyses that examine spatial relationships between variables. Related to drug therapy, GIS technology has been used to study variations in prescribing controller asthma drugs, based on clinical guidelines, for pediatric patients with asthma.[27] A GIS program was used successfully to assess the impact of a large-scale distribution program for nicotine replacement therapy in New York City.[28] In addition, GIS was proven to be a useful tool to detect local patterns of opiate drug prescribing and use. Such was the case in a study that tracked opiate prescribing and use in New Mexico.[29]

There may be a link between medication-taking behavior and neighborhood or geographic area characteristics. A new approach to identifying population or spatially associated, community-based variables may be the use of GIS technology. The first step in assessing the usefulness of GIS is to test the feasibility of mapping drug adherence in an effort to identify variations based on geography. Once the feasibility is established, further work can begin to determine the neighborhood or spatially related factors that are associated with adherence variation. Data that can be used for these analyses would include census data, population survey data such as the Behavioral Risk Factor Surveillance System available from the Centers for Disease Control and Prevention, or locally obtained information including health system infrastructure.

The purpose of this study was to determine the feasibility of using GIS technology to document geographic areas of high and low drug adherence in patients who were discharged from a large university-based hospital with continuing treatment related to ACS. To date, there are relatively few studies using GIS technology to assess drug adherence on a community or neighborhood level.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.