Evidence based? Caveat emptor!

Medical practices, clinical practice guidelines, clinical performance measures and measurements, and a variety of health care-related administrative decisions, such as insurance coverage decisions, are claiming to be "evidence based" with increasing frequency. In this paper we examine the "evidence based" label; discuss how evidence ought to have been assembled, evaluated, and synthesized; and when evidence is sufficient for the "evidence-based" moniker to rightfully apply. We also highlight several considerations other than the strength of evidence that are relevant to several common types of health care-related administrative decisions and that influence the extent to which the resulting decisions are truly evidence based.

Opportunities for improving the care of patients with chronic renal insufficiency: current practice patterns.

There are between 2 and 13 million Americans with chronic kidney disease (CKD). Recent reports suggest that their treatment is currently suboptimal. To further investigate this issue, patterns of practice for the treatment of patients with CKD who were enrolled in a large health maintenance organization in New Mexico were analyzed. Among the >200,000 patients who were enrolled in the health maintenance organization between 1994 and 1997, a cohort of 1658 patients who exhibited at least two gender-specific, elevated creatinine concentrations (Cr), separated by at least 90 d, were identified. The proportions of patients with Cr values of <2.0, 2.0 to 2.9, 3.0 to 3.9, and > or =4.0 mg/dl were 73, 17, 3, and 7%, respectively. The majority of patients were treated by a primary care physician until Cr values reached 3.0 mg/dl, at which time a nephrologist was consulted. Care tended to be transferred to the nephrologist when the Cr reached 4.0 mg/dl. Only 7.4% of patients received erythropoietin (EPO). Use of EPO increased as Cr increased. EPO was unlikely to be prescribed unless the patient had visited a nephrologist. Fewer than one half of all patients with CKD and fewer than 20% of patients with CKD with Cr values of > or =4.0 mg/dl received an angiotensin-converting enzyme inhibitor (ACEI). Nephrologists were not more likely to prescribe ACEI than were primary care physicians. Diabetic patients were more likely to receive ACEI than were nondiabetic patients, but ACEI use was quite low even among diabetic patients with CKD. The average number of hospitalizations per patient-year increased as Cr increased and was more than twice as high for patients with Cr values of > or =4.0 mg/dl, compared with those with Cr values of <2.0 mg/dl. The reasons for hospitalization were more likely to be related to comorbidities than to CKD itself, however. There are many opportunities to improve the care of patients with CKD. Better adherence to practices known to be of clinical benefit for patients with CKD not only will improve patient outcomes but also may reduce the costs of care. Providers, policy-makers, and payers should view CKD as a major public health problem and initiate innovative programs to address this growing patient population.