Jonathan Lomas

Do practice guidelines guide practice? The effect of a consensus statement on the practice of physicians.

Guidelines for medical practice can contribute to improved care only if they succeed in moving actual practice closer to the behaviors the guidelines recommend. To assess the effect of such guidelines, we surveyed hospitals and obstetricians in Ontario before and after the release of a widely distributed and nationally endorsed consensus statement recommending decreases in the use of cesarean sections. These surveys, along with discharge data from hospitals reflecting actual practice, revealed that most obstetricians (87 to 94 percent) were aware of the guidelines and that most (82.5 to 85 percent) agreed with them. Attitudes toward the use of cesarean section were congruent with the recommendations even before their release. One third of the hospitals and obstetricians reported changing their practice as a consequence of the guidelines, and obstetricians reported rates of cesarean section in women with a previous cesarean section that were significantly reduced, in keeping with the recommendations (from 72.2 percent to 61.1 percent; P less than 0.01). The surveys also showed, however, that knowledge of the content of the recommendations was poor (67 percent correct responses). Furthermore, data on actual practice after the publication of the guidelines showed that the rates of cesarean section were 15 to 49 percent higher than the rates reported by obstetricians, and they showed only a slight change from the previous upward trend. We conclude that guidelines for practice may predispose physicians to consider changing their behavior, but that unless there are other incentives or the removal of disincentives, guidelines may be unlikely to effect rapid change in actual practice. We believe that incentives should operate at the local level, although they may include system-wide economic changes.

Evaluating the message: the relationship between compliance rate and the subject of a practice guideline.

To explore the relationship between providers compliance and some key aspects of the clinical messages in practice guidelines, studies published in the English language medical literature between 1980 and 1991 were retrieved through MEDLINE and through relevant review articles in the field. All published studies providing compliance rates with practice guidelines and endorsed by official organizations were eligible for the study. The clinical content and the reported compliance rate were gathered for each recommendation in the 23 studies selected. The medical and surgical procedures addressed by 143 recommendations were identified according to specialty area, type of procedure (diagnostic, surgical, etc.) and were independently classified by the authors as being high or low on characteristics thought to influence diffusion: complexity, trialability and observability. The mean compliance rate with the 143 clinical recommendations was 54.5% (95% CI: 50.2%-58.9%), with those in the specialty areas of cardiology and oncology showing the highest compliance (mean 63.6% and 62.2%, respectively). Recommendations concerning procedures with high complexity had lower compliance rates than those low on complexity (41.9% vs. 55.9%; P =0.05), and those judged to be high on trialability had higher compliance rates than those low on trialability (55.6% vs 36.8%; P =0.03). Overall, all the characteristics of the clinical recommendations considered in the practice guidelines could account for no more than 47% of the observed variability in compliance rates. The target area of practice and the complexity and trialability of the recommended procedure appear to be useful, if partial, predictors of the level of compliance with a practice guideline.

Diffusion, dissemination, and implementation: who should do what?

The concepts of diffusion, dissemination, and implementation are distinguished as progressively more active steps in the process of flowing valid and reliable research information into clinical practice. Using a staged model of behavior change, diffusion is seen as a precursor for dissemination activities, which in turn predispose physicians to consider change in their practices. Local implementation activities capitalize on this by enabling and subsequently reinforcing the desired behavior change. Different skills are needed for each activity. Biomedical journals, with some improvements, are identified as diffusion agents. Collaboration between academics and medical organizations is best suited to the dissemination stage. Local agents, empowered by resources, are best equipped for implementation activities.

Controlling Health Expenditures- The Canadian Reality

Abstract Canada and the United States have conducted a large-scale social experiment on the effects of alternative ways of funding expenditures for health care. Two very similar societies, with (until recently) very similar systems of providing health care, have adopted radically different systems of reimbursement. The results of this experiment are of increasing interest to Americans, because the Canadian approach has avoided or solved several of the more intractable problems facing the United States. In particular, overall health expenditures have been constrained to a stable share of national income, and universality of coverage (without user charges) eliminates the problems of uncompensated care, individual burdens of catastrophic illness, and uninsured populations. The combination of cost control with universal, comprehensive coverage has surprised some American observers, who have questioned its reality, its sustainability, or both. We present a comparison of the Canadian and American data on expend...

Determinants of the increasing cesarean birth rate. Ontario data 1979 to 1982.

We analyzed the records of 454,668 births in Ontario in the years 1979 to 1982 and found that the cesarean birth rate increased from 16.5 per hundred deliveries in 1979 to 18.7 in 1982. Cesarean births were classified according to four indications: previous cesarean birth, breech presentation, dystocia, and fetal distress. The increase in the cesarean rate for each indication from 1979 to 1982 was calculated and expressed as a percentage of the total rate increase. Previous cesarean births accounted for 68 per cent of the increase, breech presentation for 14 per cent, dystocia for 4 per cent, and fetal distress for 14 per cent. The impact of previous cesarean births was indicated by an increase in the number of women presenting with this indication--from 5.8 to 7.8 per cent of all deliveries. Although the incidence of breech presentation remained stable, the cesarean birth rate increased from 54.8 to 65 per cent for this indication. There were no marked changes in either the incidence of or cesarean rates for dystocia. The incidence of fetal distress doubled (2.4 to 4.7 per cent of deliveries), but the rate of cesarean births in these cases fell from 50.5 to 32.7 per cent. These findings suggest that physicians may have begun to respond to rising professional and public concern over the increasing cesarean birth rate, but major advances in controlling this rate can be achieved only by addressing the question of vaginal deliveries for some patients who have previously had cesarean section.

Retailing Research: Increasing the Role of Evidence in Clinical Services for Childbirth

A current review of the structures and assumptions of research transfer for clinical care reveals some progress from passive diffusion to active dissemination models, but little or no progress has been made toward targeting local influences on practitioner behavior for coordinated implementation of clinically relevant research into childbirth (or other) medical practices. The implementation of scientifically valid research syntheses, such as Effective Care in Pregnancy and Childbirth (ECPC), is therefore constrained by a poorly developed marketplace for retailing research information to practitioners. A survey in Canada of the four most significant potential retailing groups demonstrated that whereas clinical and community groups were adopting the necessary knowledge and attitudes, public policy makers and administrators trailed well behind them. To increase the probability of thorough retailing of ECPC, a three-phase plan could be instituted that would identify product champions within potential retailing groups, develop implementation activities for each retailing group, and convene annual conferences.

Reluctant Rationers: Public Input to Health Care Priorities

Members of the public can adopt any one of at least three roles when providing input to public decision-making: Taxpayer, collective community decision-maker, or patient. Each of these potential roles can be mapped onto three areas of public policy decision-making in health care: Funding levels and organization for the system, the services we choose to offer under public funding, and the characteristics of those who should receive the offered services. The increasing desire to involve the public across the spectrum of health care decision-making has yet to result in a clear delineation of either which of the areas are most appropriate for public input or which of the roles we wish individual participants to adopt. The average citizen (as opposed to the self-interested patient, the provider or the manager) has so far shown little interest in contributing and rarely has the requisite skills for most of the tasks asked of him or her. The widespread motivation of governments and others for seeking public input appears to be to get the public to take or share ownership in the tough rationing choices consequent on fiscal retrenchment in health care. Evaluation of existing literature leads to the conclusion that there are only limited areas where we might wish to obtain significant public input if we adopt this widespread policy motivation. Specifically, the general public should be asked to give input to, but not determine, priorities across the broad service categories that could potentially be publicly funded. Members of the public have neither the interest nor the skills to do this at the level of specific services. The role expected of such members of the public should be made explicit and should focus on collective views of the community good rather than self-interested views of individual benefit. Groups of patients, however, should be the source of input when socio-demographic characteristics are being used to decide who should receive offered services. The role expected of these consumers is not, however, to take a self-interested perspective; rather, it is to adopt Rawls veil of ignorance to reflect compassionate views of priorities across socio-demographic characteristics. Finally, there appears to be no best method for obtaining public input that overcomes the common problems of poor information upon which to base priorities, difficulty in arriving at consensus, poor representativeness of participants, and lack of opportunity for informed discussion prior to declaring priorities. There is some suggestion, however, that panels of citizens or patients, convened on an ongoing basis and provided with the opportunity to acquire relevant information and discuss its implications prior to making consensus recommendations, offer the most promising way forward.

Aging and health care utilization: New evidence on old fallacies

The proportion of the population in the older age groups will increase dramatically over the next four decades. Furthermore, current per capita rates of hospital and medical care utilization rise sharply with age beyond the age of about 55. However, demographic trends alone do not imply health care cost increases in excess of what is supportable by normal economic growth. A cost crisis will only occur if per capita rates of utilization among the elderly increase faster than for the general population. In this paper we present some descriptive data from published sources suggesting that this has been the case over the recent past in one Canadian province. The implications for the policy debate over the effects of an aging population are discussed.

Home is where the governing is: social capital and regional health governance.

The relationship between the civic nature of a community and effective political governance by regional health boards in Canada is explored. A model is proposed that identifies components of social capital such as trust, commitment and identity, associationalism, civic participation and collaborative problem-solving. These concepts are then theoretically linked to effective governance, in particular to reflection of health needs, policy implementation, population health, fiscal responsibility and administrative efficiency. The generalizability of this model is discussed, as are current research directions and policy implications for governments. The conclusion is that governments might want to incorporate a dual perspective encompassing both the political institutions and the community structure.

Bridges between Health Care Research Evidence and Clinical Practice

Research is producing increasing amounts of important new evidence for health care, but there is a large gap between what this evidence shows can be done and the care that most patients actually receive. An important reason for this gap is the extensive processing that evidence requires before application. This article discusses a three-step model for bridging research evidence to management of clinical problems: getting the evidence straight, formulating evidence-based clinical policies, and applying evidence-based clinical policies at the right place and time. This model is purposely broad in scope and provides a framework for coordinating efforts to support evidence-based medical care. The authors purpose is to represent the roles of health informatics in the context of the roles of all the key players, including health care researchers and practitioners, health care organizations, and the public. Health informatics has already made important contributions to bridging evidence to practice, including improving evidence retrieval, evaluation, and synthesis; new evidence-based information products; and computerized aids for facilitating the use of these products during clinical decision making. However, much more innovation and coordination are needed. The authors call for health informaticians to pay balanced attention to 1) the quality of evidence embodied in information innovations, 2) the performance of technologies and systems that retrieve, prepare, disseminate, and apply evidence, and 3) the fit of information tools to the specific clinical circumstances in which evidence is to be applied. Effective interdisciplinary teams that include health services researchers and other evidence experts, clinical practitioners, informaticians, and health care managers are needed to achieve success. Informaticians can make increasingly important contributions to the transfer of health care research by joining such teams.