The influence of types of decision support on physicians' decision making.

OBJECTIVE: To determine whether physicians' post-test probability estimates are influenced by receiving test characteristics and impact their subsequent clinical decisions. DESIGN: Questionnaire based randomised controlled trial. SETTING: Mailed survey with a vignette describing an infant whose pretest likelihood of pertussis was 30% and direct fluorescent-antibody (DFA) test was negative for pertussis. SUBJECTS: Nationally representative sample of US paediatricians (n = 1502). INTERVENTIONS: Random receipt of no additional information (controls), the DFA's sensitivity and specificity (TC group) or the test's sensitivity and specificity with their definitions (TCD group). MAIN OUTCOME MEASURES: Estimated post-test probability (PTP) of pertussis, PTP of 0.50, "nearly correct" PTP (+/-5%), intended erythromycin management and intended hospital disposition. ANALYSES: Chi2 and t tests. RESULTS: Despite the negative DFA result, 67% of the 635 (49.7%) participants who responded estimated a PTP higher than the pretest probability of 30%; the overall mean estimated PTP was 0.41 (SD 0.26) (correct answer: 0.18). The TCD group's mean PTP was significantly higher than controls' mean PTP (0.45 vs 0.38, p<0.001), while the TC and control groups' mean PTP did not differ significantly (0.41 vs 0.38, p = 0.16). With decision support significantly more TC and TCD participants compared to controls estimated the PTP as 0.50 (38% vs 17%, p<0.001; 41% vs 17%, p<0.001, respectively) and also estimated a nearly correct PTP more often (20% vs 13%, p = 0.06; 19% vs 13%, p = 0.08, respectively). The mean PTP of participants intending to discontinue erythromycin therapy or discharge the patient home was significantly lower than that of participants who intended continuing erythromycin or hospitalisation (0.20 vs 0.43, p<0.001; 0.40 vs 0.49, p = 0.005, respectively). CONCLUSIONS: Paediatricians differed in their response to information about test characteristics. For many, it increased errors in estimating post-test probability; for others, it reduced errors. Estimated post-test probability was logically associated with intended clinical management.

Benchmarking and quality improvement: the Harvard Emergency Department Quality Study.

PURPOSE: To determine whether feedback of comparative information was associated with improvement in medical record and patient-based measures of quality in emergency departments. SUBJECTS AND METHODS: During 1-month study periods in 1993 and 1995, all medical records for patients who presented to five Harvard teaching hospital emergency departments with one of six selected chief complaints (abdominal pain, shortness of breath, chest pain, hand laceration, head trauma, or vaginal bleeding) were reviewed for the percent compliance with process-of-care guidelines. Patient-reported problems and patient ratings of satisfaction with emergency department care were collected from eligible patients using patient questionnaires. After reviewing benchmark information, emergency department directors designed quality improvement interventions to improve compliance with the process-of-care guidelines and improve patient-reported quality measures. RESULTS: In the preintervention period, 4,876 medical records were reviewed (99% of those eligible), 2,327 patients completed on-site questionnaires (84% of those eligible), and 1,386 patients completed 10-day follow-up questionnaires (80% of a random sample of eligible participants). In the postintervention period, 6,005 medical records were reviewed (99% of those eligible), 2,899 patients completed on-site questionnaires (84% of those eligible), and 2,326 patients completed 10-day follow-up questionnaires (80% of all baseline participants). In multivariate analyses, adjusting for age, urgency, chief complaint, and site, compliance with process-of-care guidelines increased from 55.9% (preintervention) to 60.4% (postintervention, P = 0.0001). We also found a 4% decrease (from 24% to 20%) in the rate of patient-reported problems with emergency department care (P = 0.0001). There were no significant improvements in patient ratings of satisfaction. CONCLUSION: Feedback of benchmark information and subsequent quality improvement efforts led to small, although significant, improvement in compliance with process-of-care guidelines and patient-reported measures of quality. The measures that relied on patient reports of problems with care, rather than patient ratings of satisfaction with care, seemed to be more responsive to change. These results support the value of benchmarking and collaboration.

Hospital admissions through the emergency department: does insurance status matter?

PURPOSE: To assess the effect of insurance status on the probability of admission and subsequent health status of patients presenting to emergency departments. SUBJECTS AND METHODS: We performed a prospective cohort study of patients with common medical problems at five urban, academic hospital emergency departments in Boston and Cambridge, Massachusetts. The outcome measure for the study was admission to the hospital from the emergency department and functional health status at baseline and follow-up. RESULTS: During a 1-month period, 2,562 patients younger than 65 years of age presented with either abdominal pain (52%), chest pain (19%) or shortness of breath (29%). Of the 1,368 patients eligible for questionnaire, 1,162 (85%) completed baseline questionnaires, and of these, 964 (83%) completed telephone follow-up interviews 10 days later. Fifteen percent of patients were uninsured and 34% were admitted to the hospital from the emergency department. Uninsured patients were significantly less likely than insured patients to be admitted, both when adjusting for urgency, chief complaint, age, gender and hospital (odds ratio = 0.5, 95% confidence interval 0.3 to 0.7), and when additionally adjusting for comorbid conditions, lack of a regular physician, income, employment status, education and race (odds ratio = 0.4, 95% confidence interval 0.2 to 0.8). However, there were no differences in adjusted functional health status between admitted and nonadmitted patients by insurance status, either at baseline or at 10-day follow-up. CONCLUSIONS: Uninsured patients with one of three common chief complaints appear to be less frequently admitted to the hospital than are insured patients, although health status does not appear to be affected. Whether these results reflect underutilization among uninsured patients or overutilization among insured patients remains to be determined.

The effect of supervision of residents on quality of care in five university-affiliated emergency departments.

PURPOSE: To assess the impact of direct supervision of resident physicians by attending physicians on quality of care in emergency departments. METHOD: In 1993, compliance with process-of-care guidelines was measured for 3,667 patients cared for by residents in five emergency departments in Boston and Cambridge, Massachusetts. Those patients presented with abdominal pain, asthma/COPD, chest pain, hand laceration, head trauma, or vaginal bleeding. A follow-up survey to assess patient satisfaction and reported problems with care was completed by 1,094 randomly sampled patients. RESULTS: In multivariate analysis, residents directly supervised by attending physicians had significantly (p < .0001) higher adjusted mean percentage compliance with guidelines (64%) than did residents alone (55%). Better compliance was also associated with higher level of training of the resident and greater patient urgency. There was no significant difference between supervised and unsupervised residents in either adjusted patient satisfaction or reported problems with care. CONCLUSIONS: Direct supervision of residents in emergency departments is significantly associated with better compliance with guidelines, regardless of level of training. However, direct supervision was not shown to influence patients' experience with care.

Insurance or a regular physician: which is the most powerful predictor of health care?

OBJECTIVES: This study compared the relative effects on access to health care of relationship with a regular physician and insurance status. METHODS: The subjects were 1952 nonretired, non-Medicare patients aged 18 to 64 years who presented with 1 of 6 chief complaints to 5 academic hospital emergency departments in Boston and Cambridge, Mass, during a 1-month study period in 1995. Access to care was evaluated by 3 measures: delay in seeking care for the current complaint, no physician visit in the previous year, and no emergency department visit in the previous year. RESULTS: After clinical and socioeconomic characteristics were controlled, lacking a regular physician was a stronger, more consistent predictor than insurance status of delay in seeking care (odds ratio [OR] = 1.6, 95% confidence interval [CI] = 1.2, 2.1), no physician visit [OR] = 4.5%, 95% CI = 3.3, 6.1), and no emergency department visit (OR = 1.8, 95% CI = 1.4, 2.4). For patients with a regular physician, access was no different between the uninsured and the privately insured. For privately insured patients, those with no regular physician had worse access than those with a regular physician. CONCLUSIONS: Among patients presenting to emergency departments, relationship with a regular physician is a stronger predictor than insurance status of access to care.

Relation between negligent adverse events and the outcomes of medical-malpractice litigation.

BACKGROUND: We have previously shown that in New York State the initiation of malpractice suits correlates poorly with the actual occurrence of adverse events (injuries resulting from medical treatment) and negligence. There is little information on the outcome of such lawsuits, however. To assess the ability of malpractice litigation to make accurate determinations, we studied 51 malpractice suits to identify factors that predict payment to plaintiffs. METHODS: Among malpractice claims that we reviewed independently in an earlier study, we identified 51 litigated claims and followed them over a 10-year period to determine whether the malpractice insurer had closed the case. We obtained detailed summaries of the cases from the insurers and reviewed the litigation files if the outcome of a case differed from the outcome predicted in our original review. RESULTS: Of the 51 malpractice cases, 46 had been closed as of December 31, 1995. Among these cases, 10 of 24 that we originally identified as involving no adverse event were settled for the plaintiffs (mean payment, $28,760), as were 6 of 13 cases classified as involving adverse events but no negligence (mean payment, $98,192) and 5 of 9 cases in which adverse events due to negligence were found in our assessment (mean payment, $66,944). Seven of eight claims involving permanent disability were settled for the plaintiffs (mean payment, $201,250). In a multivariate analysis, disability (permanent vs. temporary or none) was the only significant predictor of payment (P=0.03). There was no association between the occurrence of an adverse event due to negligence (P = 0.32) or an adverse event of any type (P=0.79) and payment. CONCLUSIONS: Among the malpractice claims we studied, the severity of the patient's disability, not the occurrence of an adverse event or an adverse event due to negligence, was predictive of payment to the plaintiff.