Optimal methods for guideline implementation: conclusions from Leeds Castle meeting.

BACKGROUND: Quality problems in medical care are not a new finding. Variations in medical practice as well as actual medical errors have been pointed out for many decades. The current movement to write practice guidelines to attempt to correct these deviations from recommended medical practice has not solved the problem. OBJECTIVE: In order to gain greater acceptance of these guidelines and to change the behavior of health care providers, the science of guideline implementation must be understood better. RESEARCH DESIGN: A group of experts who have studied the problem of implementation in Europe and the United States was convened. This meeting summary enumerates the implementation methods studied to date, reviews the theories of behavioral change, and makes recommendation for effecting better implementation guidelines. RESULTS: A research agenda was proposed to further our knowledge of effective evidence-based implementation.

A 2020 vision for American health care.

We enter this century with an unprecedented federal budget surplus-$4.6 trillion over the next 10 years. A substantial portion of the surplus comes from savings in the health care sector. The 1997 Balanced Budget Act cut payments to Medicare providers and raised the premiums for individual beneficiaries, but we overshot the mark. Instead of balancing the budget, we generated a huge surplus. We underestimated the magnitude of Medicare savings. Medicare savings over the period from 1998 to 2007 represent an estimated 15% of the total budget surplus. Fifteen percent of the 10-year budget surplus from 2001 to 2010 comes to $680 billion. We also underestimated the drop-off in Medicaid coverage, as welfare reform took hold. In the year 2000 Medicare and Medicaid outlays were an estimated $104 billion less than projected just 5 years ago-representing an estimated 45% of the budget surplus this year, or about $1 trillion of the 10-year surplus.

Competition on quality in managed care.

There is intense competition between managed care organizations (MCOs) in the USA based on cost and benefit coverage, but scant attention to differences in quality. Consumer preference for 'choice' has stimulated the growth of overlapping networks of providers across competing MCOs. These networks have tended to perform less well on the quality indicators in report cards than staff model MCOs. Ideally one would measure individual provider performance; but the overlapping networks, and the fact that each MCO represents a small fraction of each provider's practice, make that difficult to do. MCOs could potentially collaborate to measure individual provider performance. Financial incentives and risk-adjusted premiums might stimulate competition on quality within MCOs. It seems more likely that true competition on quality will occur between groups of providers, organized or integrated delivery systems, than between MCOs. Nevertheless, MCOs are likely to offer some quality-improving programs directly to their members, and can stimulate the competition between providers by collaborating to obtain provider-specific measurements.

Postdischarge follow-up of psychiatric inpatients and readmission in an HMO setting.

This study examined the relationship between follow-up and rehospitalization of inpatients discharged from treatment in two divisions of a health maintenance organization (HMO). Among 580 patients discharged, two-thirds made a follow-up visit within 30 days. Slightly less than a third were readmitted within six months. Readmission was less likely for patients who made a follow-up visit and for men. Patients who had a preadmission relationship with a mental health practitioner were more likely to make a follow-up visit and were more likely to be readmitted. Follow-up was also associated with diagnoses of adjustment and affective disorders.

Relationship of anatomic disease to appropriateness ratings of coronary angiography.

BACKGROUND: At Harvard Community Health Plan (HCHP), Brookline, Mass, a mixed-model health maintenance organization (HMO), coronary angiography is performed at numerous community and tertiary-level teaching hospitals. OBJECTIVE: To determine the appropriateness of coronary angiography within HCHP according to RAND (1992) criteria and to examine the relationship between the appropriateness rating and (1) the clinical indication for catheterization and (2) the extent of anatomic disease. METHOD: A retrospective, randomized hospital medical record review of 292 patients enrolled in HCHP who underwent coronary angiography in 1992, stratified by four distinct HCHP subgroups. RESULTS: Of the coronary angiographies reviewed, 78% were rated appropriate, 16% uncertain, and only 6% inappropriate across the entire sample. Ratings were comparable in all subdivisions of HCHP despite an incidence rate of catheterization in one of the three HMO divisions that was 60% and 40% higher than in the other two divisions. The lowest appropriateness ratings were for Asymptomatic patients (43%) and those with Chest Pain of Uncertain Origin (35%) (capital letters refer to the RAND clinical indication criteria mentioned above). A rating of necessity was not a better discriminator of anatomic disease than a rating of appropriateness alone: 82% and 84%, respectively, were found to have disease by angiography. CONCLUSION: The low HCHP rate of inappropriateness for coronary angiography is comparable with the RAND 1992 New York State data. This finding, coupled with marked differences in the incidence rate of this procedure among the HCHP divisions, is consistent with either major differences in the sickness of the HMO's sub-populations or, more likely, a lack of specificity of the RAND criteria for coronary angiography.

The "teaching HMO": a new academic partner.

Health care reform is a potential threat to the academic missions of medical schools and academic health centers. But managed care, the source of much of their concern, may also represent a way for medical schools to improve their future academic outcomes. Harvard Medical School and the Harvard Community Health Plan, a large health maintenance organization (HMO) in greater Boston, recently formed the first medical school department to be based in a freestanding HMO. This arrangement is an example of a model that replicates, in a managed care organization, the long-standing and highly successful teaching hospital academic structure in academic medical centers. The authors describe this model in detail, show how the Harvard collaboration works, and explain the benefits each institution saw in creating a joint entity, the rationale for making that new entity an academic department, and the implications for other academic health centers. They conclude that the Harvard experience shows that alliances between medical schools and large HMOs can create vibrant practice settings for teaching and research in academic areas (such as prevention and primary care medicine) that have been relatively neglected in recent times, and that the "teaching HMO" may have the potential to transform academic medicine in the next century just as the teaching hospital transformed it in this century.

ABO and rhesus blood groups and adverse outcomes of pregnancy.

Interview and medical-record data of 11,659 nondiabetic, non-asthmatic women were analyzed to evaluate the relationship between ABO and Rh blood groups and adverse outcomes of pregnancy. No statistically significant associations were found in logistic regression analyses that controlled for age, race, smoking and parity. The results of our study demonstrate the importance of controlling for confounding variables, and do not support a relationship between maternal blood group and adverse pregnancy outcomes including malformations.

What's ahead on quality: the managed care perspective.

Without question, the most important processes occurring in managed care that can be expected to affect quality are accreditation and the effort to obtain and compare uniform information on quality of care across health care organizations, in short, to create "report cards." For both processes, 1993 was an extremely productive year, and 1994 promises to be even more so. These two processes fit hand-in-glove--one is designed to determine that managed care organizations are equipped to serve the public and to implement better health care programs, while the other is designed to help them understand and improve their own performance. Although, in the short run, managed care organizations may view both these efforts as additional costs, in the long run, both should lead to a better industry and to better care for the public.

Feedback of clinical performance information.

Measurement of practice performance, increasing throughout the health care industry, can result not just in performance assessment, but in performance improvement. An important early step in achieving improvement is to feed back the performance information. This paper discusses how the content of the information and the process of the feedback itself can facilitate physicians' buying into and using the data to improve practice.

Implementing clinical guidelines: a quality management approach to reminder systems.

"Reminders" (timely notices about specific clinical events to physicians or patients) are useful strategies for implementing clinical guidelines. These systems can vary widely in content, design, and purpose. This variability makes it difficult to compare systems or predict a reminder's efficacy in a particular setting. In this article, the authors suggest that too much attention has been paid to proving the efficacy of reminder systems as a general strategy or a comprehensive solution. Rather, the usefulness and effectiveness of reminder systems will be better appreciated when quality management principles are applied to designing, implementing, and maintaining them. Physicians must participate in finding the best solutions for clinical process problems, and reminder systems--when implemented as the "best solution"--must be monitored and improved continually on an ongoing basis to continue to meet physicians' needs.

Cigarettes, coffee, and preterm premature rupture of the membranes.

Premature (prior to 37 completed weeks of gestation) rupture of the membranes (preterm PROM) is one of the most common underlying causes of preterm delivery. However, there have been few epidemiologic studies of this obstetric complication. The authors studied the relation of maternal cigarette smoking and coffee consumption to both preterm PROM and spontaneous preterm labor not complicated by premature rupture of the membranes (preterm NONPROM) in a large cross-sectional data base. The 307 preterm PROM and 488 preterm NONPROM cases who delivered during 1977-1980 at the Boston Hospital for Women were compared with 2,252 randomly selected women who delivered at term at that institution. Multiple logistic regression techniques were used to derive maximum likelihood estimates of adjusted odds ratios (OR) and 95% confidence intervals (CI). After confounders had been adjusted for, the relative risk of preterm PROM for women who reported ever having smoked during pregnancy, as compared with nonsmokers, was 1.6 (95% CI 1.1-2.4). However, no gradient between the number of cigarettes smoked per day and the risk of preterm PROM was observed. Similar results were observed for preterm NONPROM. Women who consumed three or more cups of coffee daily during the first trimester had a 2.2-fold greater risk of preterm PROM than did women who drank two or fewer cups (95% CI 1.5-3.3). Among coffee drinkers, there was some evidence of a linear trend in the risk of preterm PROM as coffee consumption increased. Consumption of three or more cups of coffee per day was less strongly associated with the occurrence of preterm NONPROM (adjusted OR = 1.4, 95% CI 1.0-1.9).

Algorithm-based clinical quality improvement. Clinical guidelines and continuous quality improvement.

Evidence documenting unexplained variation in clinical practices and outcomes has led to a proliferation of clinical practice guidelines in the hope that such efforts will lead to decreased variation, improved care, better outcomes, and lower costs. At Harvard Community Health Plan we have developed a clinical guideline development effort that focuses on the development of clinical algorithms and guidelines in a quality improvement model. The formal quality improvement process that we have described requires; (1) clear project definition and organization, (2) guideline development based on an understanding of patient needs, scientific evidence and clinical experience, (3) thorough analysis of potential problems with active implementation efforts, and (4) measurement and evaluation of results. By incorporating clinical guideline development into a quality improvement model and integrating such efforts with a total quality management strategy, we can substantially increase the likelihood of successfully implementing clinical practice guidelines and improving the quality of care that we deliver.

Automated ambulatory medical records systems. An orphan technology.

Automated ambulatory medical records systems (AAMRSs) have been operational for over 20 years but have not been adopted by more than a small fraction of their potential users. This paper presents a detailed analysis of the uses and benefits of the COSTAR-based AAMRS at the Harvard Community Health Plan and of the factors which have inhibited the dissemination of COSTAR. We conclude that AAMRSs have been an orphan technology and cite trends in health care that favor the future development of AAMRSs.

Risk factors for abruptio placentae.

The authors studied the relation of sociodemographic, medical, and life-style factors and abruptio placentae in a large cross-sectional data base. The 143 cases of abruption identified in the Delivery Interview Program, conducted in Boston from 1977 to 1980, were compared with 1,257 randomly selected controls. The authors used multiple logistic regression techniques to derive maximum likelihood estimates of the adjusted odds ratios and 95% confidence intervals as measures of the association between exposure factors and abruption. A history of chronic hypertension was associated with a threefold increase in risk (odds ratio (OR) = 3.1, 95% confidence interval (Cl) 1.1-8.4), but pregnancy-induced hypertension was not associated with abruption. Excess risk was associated with advanced maternal age (OR = 2.3, 95% Cl 1.3-3.9), low prepregnancy body mass index (OR = 2.3, 95% Cl 1.3-4.1), a history of prior stillbirth (OR = 3.5, 95% Cl 1.8-7.0), and at least weekly use of marijuana during pregnancy (OR = 2.8, 95% Cl 1.2-6.6). Overall, the association with cigarette smoking during pregnancy was of only borderline significance (OR = 1.5, 95% Cl 1.0-2.2), although there appeared to be a dose-response relation between the number of cigarettes smoked and risk of abruption.

Cigarette smoking during pregnancy in relation to placenta previa.

To evaluate the relationship between cigarette smoking and the occurrence of placenta previa, we used interview and medical record data to conduct a case-control analysis of 69 placenta previa cases and 12,351 controls. The unadjusted relative risk estimate of placenta previa for women reported to have "ever smoked" during pregnancy relative to nonsmoking mothers was 1.9 (95% confidence interval, 1.2 to 3.0). The risk rose after adjusting for potential confounders (odds ratio, 2.6; 95% confidence interval, 1.3 to 5.5). In contrast to a previous report, the duration of smoking was not an independent risk factor for placenta previa. These results suggest that cigarette smoking during pregnancy is a determinant of placenta previa. Carbon monoxide hypoxemia, which is one possible mechanism for this association, may result in compensatory placental hypertrophy. Placentas with increased surface areas are more likely to cover the cervical os, causing placenta previa.