Douglas K. Martin

The Regenstrief Medical Record System: a quarter century experience

Entrusted with the records for more than 1.5 million patients, the Regenstrief Medical Record System (RMRS) has evolved into a fast and comprehensive data repository used extensively at three hospitals on the Indiana University Medical Center campus and more than 30 Indianapolis clinics. The RMRS routinely captures laboratory results, narrative reports, orders, medications, radiology reports, registration information, nursing assessments, vital signs, EKGs and other clinical data. In this paper, we describe the RMRS data model, file structures and architecture, as well as recent necessary changes to these as we coordinate a collaborative effort among all major Indianapolis hospital systems, improving patient care by capturing city-wide laboratory and encounter data. We believe that our success represents persistent efforts to build interfaces directly to multiple independent instruments and other data collection systems, using medical standards such as HL7, LOINC, and DICOM. Inpatient and outpatient order entry systems, instruments for visit notes and on-line questionnaires that replace hardcopy forms, and intelligent use of coded data entry supplement the RMRS. Physicians happily enter orders, problems, allergies, visit notes, and discharge summaries into our locally developed Gopher order entry system, as we provide them with convenient output forms, choice lists, defaults, templates, reminders, drug interaction information, charge information, and on-line articles and textbooks. To prepare for the future, we have begun wrapping our system in Web browser technology, testing voice dictation and understanding, and employing wireless technology.

Computerized Display of Past Test Results: Effect on Outpatient Testing

STUDY OBJECTIVE To determine the effect of displaying previous results of diagnostic tests on the ordering of selected outpatient tests. DESIGN Sixteen-week controlled trial with a 13-week pre-intervention and 8-week post-intervention observation periods. Patients were randomly assigned to intervention or control groups so that each physician was his or her own control. Only scheduled visits were included. Randomization occurred before the pre-intervention observation period. SETTING Academic primary care general medicine clinic affiliated with an urban general hospital. SUBJECTS Pre-intervention period: 111 physicians (97 internal medicine residents, 14 faculty internists), 4683 patients, 5942 scheduled visits. INTERVENTION period: same 111 physicians, 5946 patients, 8148 visits. Post-intervention period: 76 physicians (62 residents, 14 faculty), 2571 patients, 2858 scheduled visits. INTERVENTION With an order for one of eight selected diagnostic tests through microcomputer workstations, a window was opened on the screen and previous test results were displayed along with the time interval between the first and last result. Tests were ordered for control patients into the same workstations without previous results displayed. MEASUREMENTS AND MAIN RESULTS Previous results of one or more study tests were available for display for 96% of scheduled patients. Significantly lower results (p less than 0.05 by paired t-test) for the selected tests were found for intervention patient visits than for control visits: charges per visit (mean +/- SE) for intervention patients

The prognosis of hyponatremia at hospital admission

12.17 +/- 0.62, compared with

Examination of Racial Differences in Management of Cardiovascular Disease

13.99 +/- 0.77 for controls, a 13.0% difference; tests per visit were 0.51 +/- 0.03, compared with 0.56 +/- 0.03, an 8.5% difference. The number of study tests ordered decreased significantly for intervention patients (16.8%) and for controls (10.9%). During the post-intervention period, ordering of study tests increased for both groups, but the increase from the intervention period was not significant. CONCLUSIONS Presenting physicians with previous test results reduced the ordering of those tests. The actual effect may have been greater than 13%, because there were reductions in study tests ordered for both intervention and control patients during the intervention period when compared with the pre-intervention period, and both tended to rise after the intervention, or display, was turned off.

Primary care physician productivity - The physician factor

To assess the risk of mortality in patients with hyponatremia at the time of hospital admission, the authors studied data for 13,979 patients admitted over a 46-month period. Of the 763 (4%) admitted with hyponatremia, 757 (99%) were matched by age, gender, and admitting date with normonatremic control patients. Hyponatremic patients were more than seven times as likely to die in the hospital than the control patients, and they were more than twice as likely to die after discharge (p<0.0001 for both). This relationship with in- and outpatient mortality held when controlling for the diagnoses found more often in the hyponatremic patients. Hyponatremia appears to be an indicator of increased risk of death regardless of the disease with which it is associated.

Using clinical data to predict abnormal serum electrolytes and blood cell profiles

OBJECTIVES We sought to identify the clinical characteristics associated with, and to investigate the impact of cohort selection criteria on, interracial use of invasive cardiac procedures and to determine survival. BACKGROUND Although interracial differences in the use of invasive cardiac procedures have been previously reported, the underlying reasons are not known. METHODS A retrospective cohort study was conducted at a Veterans Affairs Medical Center. Study patients were evaluated for cardiovascular disease between January 1 and December 31, 1993. RESULTS The study included 1,406 male patients (85% white, 58% married), with a mean age of 63.4 years. African Americans were less likely than whites to undergo procedures (cardiac catheterization: odds ratio [OR] 0.37, 95% confidence interval [CI] 0.24 to 0.58; coronary angioplasty: OR 0.60, 95% CI 0.25 to 1.49; coronary bypass surgery: OR 0.22, 95% CI 0.08 to 0.63; any procedure: OR 0.32, 95% CI 0.21 to 0.50). On bivariate analysis, patients who underwent cardiac procedures were more likely to be younger, married and reside nonlocally and less likely to have severe comorbid disease; however, African Americans were less likely to be married and to reside nonlocally and more likely to have severe comorbid disease. Cohorts adjusting for referral status and specified cardiac diagnoses reduced or reversed interracial treatment differences. Thirty-day and 1-year survival rates (96% and 87.6%, respectively) were equivalent. CONCLUSIONS Racial disparity in invasive cardiac procedure use may be partially explained by clinical differences and cohort selection bias. Despite treatment differences, survival rates were equivalent in African Americans and whites.

The value of serum iron studies as a test for iron-deficiency anemia in a county hospital

AbstractOBJECTIVES: To model physician productivity as a function of clinic (support system) characteristics and physician characteristics and to model the time a physician spends with the patient as a function of patient characteristics. DESIGN: Observational study. SETTING: A general medicine clinic of a university-affiliated Veterans Affairs medical center. PATIENTS: A cohort of 2,520 patients having 2,721 consecutive outpatient visits to 56 physicians. MAIN OUTCOME MEASURES: Physician productivity defined as patients seen/physician/hour and time (minutes) spent with the patient. RESULTS: Physicians saw a mean (±SD) of 1.62±0.68 patients/hour. Clinic characteristics explained 8.2% of the variability of session-specific physician productivity. Controlling for clinic characteristics, a factor representing the physician explained an additional 55.4%. A model for overall physician productivity, using physician characteristics, explained 84.9% of the variance, and time spent with the patient was an important predictor. Modeling physician time with patients, patient characteristics accounted for only 7% of the variability. Controlling for patient characteristics, the individual physician again provided the greatest explanatory power, an additional 22.8% of the variability. CONCLUSIONS: Physicians’ practice patterns, rather than clinic or patient characteristics, may account for most of the variation in physician productivity. Given the magnitude of the influence of individual practice patterns, interventions to increase productivity need to consider methods to affect physician behavior.

SOME TECHNIQUES FOR THE ANALYSIS OF WORK SAMPLING DATA

Objective: To identify clinical predictors of five abnormalities on the serum electrolyte panel and two abnormalities on the blood cell profile, to study which data elements carried predictive information, and to measure the predictive accuracy and stability of the resulting predictive equations.Design: Prospective data collection from a computerized medical database supplemented by data entered by physicians who ordered outpatient tests into microcomputers. Equations were derived during an eight-month period and later validated twice in the same setting.Setting: Academic primary care practice affiliated with a county hospital.Patients and participants: Patients were mostly black women; physicians were full-time academic general internists and medical residents.Measurements and main results: There were 6,570 electrolyte and blood cell profile panels ordered during the equation derivation period. The mean receiver operating characteristic (ROC) curve area for the seven equations was 0.849. For the 4,977 tests ordered during ten months of prospective validation, the mean ROC curve area was only 3% less. For three equations, ROC curve areas were lower for patients with unscheduled visits than for those with scheduled visits (p<0.05). Except for two equations involving abnormalities with very low prevalences, the equations were also well calibrated. Prior results for the abnormality being considered were the strongest predictors, followed by other laboratory results, diagnoses, and the physicians’ estimate of the probability that the test would be abnormal.Conclusions: Clinical data can accurately predict abnormal results of common outpatient laboratory tests. Computers can help find the necessary data and produce estimates of risk.

Computer Predictions of Abnormal Test Results: Effects on Outpatient Testing

To evaluate the performance of serum iron studies as a diagnostic test for iron-deficiency anemia in a county hospital, the authors identified retrospectively all general medicine patients who had had bone-marrow aspirates for the work-up of non-macrocytic anemias from 1978 through 1983. Re-reading a sample of aspirates from the 254 study patients (42 with iron deficiency) verified the presence of absence of iron. Analysis with logistic regression, likelihood ratios, and receiver operating characteristic curves demonstrated that the total iron-binding capacity (TIBC) performed markedly better as a diagnostic test than did the transferrin saturation test. While no single TIBC level was diagnostic, the TIBC provided a good estimate of the probability of iron-deficiency anemia.

Racial disparity in cardiac decision making: results from patient focus groups.

Work sampling is an observational technique that produces counts representing the number of times that an individual has been observed performing each of several tasks. These data are collected using either systematic or random times of observation, and typically exhibit correlation between repeated observations on the same individual, with the degree of correlation being a function of the amount of time elapsed between measurements. Using several recently developed statistical techniques, we illustrate how it is possible to carry out analyses of these nominal outcomes that account for the correlation between repeated outcomes. We use description of a work sampling study to motivate the techniques and we compare empirically results from analyses based on several different underlying assumptions.