A Summary Of the Population Health Management Journal

The Disease Management Care Blog's travels have been taking it back and forth to the Population Health and Care Coordination Colloquium. It's been taking copious notes at that meeting (more on that in a future post), but the travel-down time has been a DMCB opportunity to curl up and take notes on the articles in the latest issue of the Population Health Management Journal.

Brief article summaries posted below.

Yi Zhou, Robert Unitan, Jian Wang, Terhilda Garrido, Homer Chin, Marianne Turley, Linda Radler: Improving Population Care with an Integrated Electronic Panel Support Tool

In this study, the authors set out to study the impact of an electronic health record “integrated technology application” called the "Panel Support Tool" at Kaiser Northwest in Oregon and Washington state. The “PST” provided on screen point-of-care reminders, supported a patient registry and generated performance feedback using individual patient screen shots and practice panel summaries. Physicians were subjected to lab, medication and screening “care gap” reminders based on national guidelines, HEDIS measures and organizational priorities. To perform the study, the authors measured a series of cross sectional “care performance percentages” that consisted of the number of completed care recommendations (the numerator) divided by the number of indications for a total of 13 quality measures. These were obtained every 4 months for 20 months for 207 practice teams that collectively cared for over 263,000 patients. At baseline, the roll-up measure was 72.9% and it gradually increased to 80% at 20 months. During the study, there were no other "care gap" interventions, so even though this is pre-post, the authors believe the PST was responsible for the 7% improvement. Hardly a breakthrough, says the DMCB. But then again, nothing is.

Dahlia Remler, Jeanne Teresi, Ruth Weinstock, Mildred Ramirez, Joseph Eimicke, Stephanie Silver Steven Shea. Health Care Utilization and Self-Care Behaviors of Medicare Beneficiaries with Diabetes: Comparison of National and Ethnically Diverse Underserved Populations.

The authors used the "IDEATel" telephone survey to compare urban versus rural diabetes care for persons living in two medically underserved areas. 755 respondents (75% and 28% self identified themselves as black or latino) from northern Manhattan or the Bronx were compared to 867 respondents in rural upstate New York (35% and 17% reported black or latino). To be included, respondents had to be Medicare beneficiaries, be age greater than 55 years, diagnosed with diabetes and live in a “federally designated medically underserved” or a “federally designated health professional shortage area” in New York State. The survey was conducted from December 2000 to April 2003 and took respondents approximately ½ hour to complete. Compared to the rural group, the urban group had worse general health (11.6% vs. 5.1% self-rated their health as “poor”), more inpatient days (a mean of 3.48 vs. 1.53), more ER visits and more difficulty with a variety of self-care activities. The DMCB finds it difficult to generalize these data to the rest of the U.S. and wonders if the urban vs. rural differences were the result of other sources of unidentified bias.

Donald Fetterolf: Long-Term Results Evaluation in Medical Management Programs.

Here’s a solid review paper that tells you everything you need to know about the evaluation of the long term impact of disease management programs. This includes “bending the trend,” the phenomenon of a therapeutic “plateau” once any interventions have reached a steady state and the occurrence of a “step function,” which signals the impact of a one-times savings impact. While you may think savings between programs should be easy to calculate, that’s not the case, thanks to a lack of standard methods, groups not being comparable, a lack of control groups, evolving practice standards and shifting program content. In addition, cost is not the only consideration. There is clinical impact, quality of life, future risk pool effects, branding and employee retention. Once a program has plateaued, it’s tempting to turn it off, risking an upward “rebound” in cost trending. Once again, Dr. Fetterolf takes a complicated topic and puts it within reach of us mere mortals.

Patricia Harrison, Pamela Hara, James Pope, Michelle Young, Elizabeth Rula: The Impact of Postdischarge Telephonic Follow-up on Hospital Readmissions.

This is a study involving 30,272 Medicare Advantage plan members who were automatically enrolled in a chronic disease management program and who had a hospital admission for any reason in calendar year 2008. Any members who were discharged from a hospital were subjected to a “Hospital Discharge Campaign” that notified the provider of discharge, called the patient with 14 days of discharge, reviewed the orders to delete any duplicated or contraindicated prescriptions and made sure the patient understood “proper steps to take,” like participating in a timely PCP follow-up visit. The intervention study group numbered 6773 patients. 23,499 patients who were re-admitted prior to getting the call or who did not get the telephone call comprised the convenience comparison group. In general, being older, male and having an increased length of stay was associated with a 30 day readmission. Admissions were highest 2-3 days after discharge; a third of all admissions occurred within 7 days. While patients who got the call were 23% less likely to be readmitted, the DMCB is concerned about the use of a control group that may not be a good comparator. The observed difference could be due to factors other than the phone call. This is a good start, but better and more research is needed.

Ron Cantrell, Julie Priest, Christopher Cook, Jack Fincham Steven Burch: Adherence to Treatment Guidelines and Therapeutic Regimens: A US Claims-Based Benchmark of a Commercial Population.

This was a “cross sectional and retrospective” study of HEDIS-like and national guidelines-based quality of care measures using pooled commercial insurer claims from the Ingenix Impact National Managed Care Benchmark Database. This contains information from a whopping 45 managed care organizations with 65 million enrollees. The authors focused on 2007 claims with persons who had identifying claims in 2006 (except for new depression) or 2007 consistent with asthma, COPD, CAD, depression, diabetes, heart failure, hyperlipidemia or hypertension. Based on the sample of approximately 5.5 million people with an index condition who had at least 6 months of enrollment the authors discovered most people with asthma don’t refill their inhalers, 3% of persons with COPD will have an exacerbation, the majority of persons with new depression fail to refill their antidepressants, and 44% and 48% of persons with diabetes don’t have an A1c or lipid test, respectively. There is lots of other information here, including tables of the average number of hospitalizations, ER visits and mean claims expense. Shortcomings of the study were listed by the authors and included the limited accuracy of claims, the lack of risk adjustment between MCOs and the potential non-generalizability of the time span that was used. The DMCB didn’t find anything that was surprising, but this paper could serve as a useful benchmark for comparison purposes.

Julie Priest, Ron Cantrell, Jack Fincham, Christopher Cook, Steven Burch: Quality of Care Associated with Common Chronic Disease in a 9 State Medicaid Population Utilizing Claims Data: An Evaluation of Medication and Health Care Use and Costs.

This is the Medicaid companion to the commercial study mentioned above. It too was a “cross sectional and retrospective” study, but this time it used pooled fee-for-service Medicaid claims from nine states using the Thomson Reuters Market Scan Multi-State Medicaid data base for asthma, COPD, coronary artery disease, depression diabetes, heart failure, hyperlipidemia and hypertension. Once again, the authors examined 2007 claims on persons identified with the index condition in 2006 or 2007; dual eligible and persons older than age 65 were excluded. HEDIS-like and national guidelines-based quality of care metrics were used. There were 2.8 million individuals with an average age of about 16 years. Only 23% regularly refilled their prescriptions on time and only about a third of newly diagnosed depression patients filled any prescriptions for an antidepressant. 70% and 63%, respectively, of persons with diabetes had A1c and LDL testing. There is other information, including hospitalizations, ER visits and means claims expense. Once again there was nothing surprising, but the numerous tables could serve as a benchmark.

Image from Wikipedia