Modeling therapeutic strategies in rheumatoid arthritis: use of decision analysis and Markov models.

OBJECTIVE: The management of patients with rheumatoid arthritis (RA) is controversial, with a number of different proposed treatment strategies based on different conceptions of the natural history of the disease and different interpretations of the efficacy and effectiveness of the drugs used for treatment. We attempted to develop a theoretical framework to assess the effectiveness of different treatment regimens for RA. METHODS: We used decision analysis to structure the problem of comparing sequential monotherapy to a combination strategy. Subsequently, we used 3 different estimates of drug effectiveness: one from expert rheumatologists; a metaanalysis; and a recent nationwide survey of American rheumatologists, in a Markov model. Last, we utilized published duration of therapy data to model drug treatment over time. RESULTS: Estimates of drug effectiveness differed substantially among rheumatologists, but regardless of the estimates and the treatment strategy used, the model predicted over 90% of patients improved by the 3rd drug trial. Over time, treatment patterns in our model resemble the "sawtooth" pattern previously observed. CONCLUSION: Treatment strategies in RA are difficult to model because of uncertainty in both the structure of the model and the data needed to perform the analysis. These models tend to overestimate the effectiveness of drug sequences because of nonindependence between therapies, probably due to sequence effects, a change in responsiveness over time, or resistant subgroups. Our preliminary analysis suggests that the most effective agent, possibly methotrexate, should be used first if the objective is to get as many patients into remission as quickly as possible.

Sealant use in public and private insurance programs.

Dental sealants are a safe and efficacious method for preventing caries in pits and fissures. An increase in sealant use has been observed in the NHANES III survey. However this increase still accounts for less than one fifth of children aged 5-17 having sealants on their teeth. Reimbursement and coverage of sealants by public and private insurance programs have lagged behind the widespread acceptance of this methodology by the profession. Medicaid preventive services are poorly used by patients and the profession: and the inclusion of sealants in traditional, indemnity, fee-for-service commercial plans has been inconsistent. Managed care programs include dental sealants, but lack financial incentives to increase their use.

Health services research and systemic lupus erythematosus: a reciprocal relationship.

Many other statistical and epidemiologic approaches have been used in studies of systemic lupus erythematosus. Many of these, such as multivariate analysis and cost-benefit analysis, have substantially added to our understanding of the disease. However, in these instances it has been a clear application of the technique to a clinical problem. In the four areas I described, there was a convergence of clinical dilemmas and methodologic improvements, with the clinical problem actually contributing to the development of the methodologic technique used to address the question. In conclusion, major methodologic advancements in health services research and clinical epidemiology have developed pari passu with studies of the natural history of systemic lupus erythematosus. In some circumstances the clinical questions drove methodologic innovation and in other cases the methodologic studies were rapidly adopted and adapted to clinical investigation. It is unlikely that this was happenstance. We can anticipate that future investigations in lupus will utilize innovative techniques in health services research.

The diagnosis of polyarteritis nodosa. I. A literature-based decision analysis approach.

We investigated diagnostic testing in polyarteritis nodosa (PAN) by calculating, from published data, the sensitivity and specificity of visceral angiography and muscle, nerve, testicle, kidney, and liver biopsy. Test sequence strategies were constructed by Bayesian inference using a computer program written for this purpose. Test sequences were compared with an aggressive strategy consisting of repeated tests until there was a positive finding or until the available tests were exhausted, and a conservative strategy consisting of 1 biopsy procedure plus angiography. The Bayesian analysis agreed most closely with the conservative approach for most prior probabilities (degree of suspicion) that a patient had PAN. The aggressive strategy had an overall sensitivity of 90% and specificity of 91%, whereas the conservative strategy was 85% sensitive and 96% specific. Furthermore, the aggressive strategy was more costly ($2,986 versus $1,961) and had a higher rate of morbidity (3.8 versus 2.7 days of hospitalization per patient evaluated) than did the conservative strategy. The mortality rates of both strategies were equivalent (approximately 0.05 deaths per hundred patients evaluated). The per-case cost of diagnosis increased as prevalence decreased, and at 10% prevalence, the aggressive strategy cost more than $17,000 per case diagnosed. Sensitivity analysis revealed that the strategies were moderately affected by the test characteristics, within reasonable assumptions, but that the differences in conservative and aggressive approaches remained. Thus, our analysis based on available data and the assumption of test independence suggests that the preferred diagnostic evaluation of patients with symptoms suggestive of PAN consists, in most cases, of a single biopsy procedure, with angiographic evaluation if necessary.

The diagnosis of polyarteritis nodosa. II. Empirical verification of a decision analysis model.

We evaluated our literature-based estimates of diagnostic test characteristics and aggressive and conservative strategies for the diagnosis of polyarteritis nodosa (PAN) by reviewing 1980 through 1985 data from the University of Chicago Medical Center, Michael Reese Hospital, and Northwestern University Medical School. Test specificity was calculated by reviewing pathology and radiology reports on all relevant procedures done at the University of Chicago Medical Center in 1984 and 1985. There were no reports of false-positive findings (including angiography); thus, test specificity was 100% for muscle, nerve, kidney, liver, and testicular biopsy, which was comparable with our literature-based estimate of 97%. Test sensitivity was based on the 18 confirmed cases of PAN from the 3 institutions and was similar to that in published reports, ranging from 0% for liver biopsy to 100% for visceral angiography. Review of each case for diagnostic test sequence showed an average of 2 diagnostic procedures to confirm the diagnosis (range 1-6). Eight of 18 patients were evaluated according to the conservative strategy we proposed from our literature-based decision analysis approach. No patient was evaluated with the aggressive strategy, although 1 patient had 6 invasive procedures. Of the remaining 10 patients, 9 represented cases that might have been confirmed had the conservative approach been used, and if it had been used, 6 patients would have had the diagnosis confirmed. Thus, the minimum sensitivity of our conservative strategy is 78%, although based on these data, it could be as high as 100%.(ABSTRACT TRUNCATED AT 250 WORDS)

Prognosis in SLE: comparison of Markov model to life table analysis.

A Markov model of prognosis was evaluated by comparing the duration of disease activity states and life expectancy with Kaplan-Meier survival curves for 98 patients with systemic lupus erythematosus with 1080 patient-years of observation. A four state (remission, active, flare and death) homogeneous Markov chain was constructed to determine the transition probabilities between disease states, the probability of disease states over the subsequent 25 years, time in each disease state, and life expectancy. Approximately 85% of the clinical course of patients was in the active disease state; no patient made a transition directly between remission and flare; virtually all deaths occurred from the flare disease state. The proportion of the patient population in each disease state over time provided a convenient graphic summary of the natural history of SLE that supplemented the Kaplan-Meier survival curves. A Markov model analysis yields a clinically useful description of outcome for multistate diseases that supplements survival curves.

Decision theory in medicine: a review and critique.

Decision theory's role in medicine will lie between the extremes of naive optimism ("a Rosetta stone") and unmitigated pessimism ("a computerized Ouija board"). Its application to public health policy, research, and administrative problem-solving is established; experience offers some guidelines for more limited use in clinical practice.