Publications

Medical technology assessment has been proposed as a way to encourage more appropriate use of medical technologies, which may in turn lower medical care costs. The application to a diagnostic technology of techniques for medical technology assessment is discussed, using upper gastrointestinal endoscopy as an example. Several generic problems are often encountered when performing an assessment of a diagnostic technology. These problems include the need to weigh relatively concrete data on clinical and economic costs against less quantifiable data on the value of the information gained from the procedure. In the case of upper gastrointestinal endoscopy, data on morbidity (approximately two cases per 1,000 cases), mortality (approximately one case per 20,000 cases), changes (approximately +290 per procedure), and costs (between +69 and +128 per procedure) are relatively easily determined. Less easily calculated is the marginal diagnostic gain from endoscopies performed for the wide variety of conditions for which an endoscopy may be indicated. It is concluded that, in the case of diagnostic technologies, technology evaluations may be most useful as a heuristic tool. Because of the difficulty in weighing costs and benefits, however, formal evaluations of diagnostic technologies are not likely in the near future to contribute to medical care cost containment.

We undertook a study to determine whether moderate exercise modifies the bronchoconstriction produced by sulfur dioxide (SO2) in subjects with mild asthma. In 7 subjects, we compared the changes in specific airway resistance (SRaw) produced by 10 min of exercise alone (400 kpm/min on a cycle ergometer), inhalation of SO2 alone, and the combination of exercise and SO2. During all studies, a subject breathed SO2 and/or air from a mouthpiece. In 6 additional subjects, we compared the increase in SRaw produced by inhalation of SO2 during exercise with that produced by eucapnic hyperventilation with SO2. Neither inhalation of 0.05 ppm of SO2 at rest nor exercise or hyperventilation alone had any effect on SRaw. Inhalation of SO2 during exercise, however, significantly increased SRaw (from 8.46 +/- 3.58 L x cm H2O/L/s (mean +/- SD) to 18.16 +/- 10.05 at 0.05 ppm and from 8.07 +/- 2.69 to 10.48 +/- 4.49 at 0.25 ppm (p less than 0.05)). In the 2 most responsive subjects, inhalation of 0.10 ppm of SO2 during exercise also significantly increased SRaw. The SRaw increased by the same amount whether SO2 was inhaled during exercise or during eucapnic hyperventilation at the same minute ventilation, but the time course of the increase in SRaw was different. The SRaw was at or near maximal values at the first measurement (30 s) after hyperventilation but not until 2 to 4 min after exercise. When 4 subjects took larger breaths after inhaling SO2 during eucapnic hyperventilation to more closely match the volume of the breaths taken after exercise, the time courses of SO2-induced bronchoconstriction after hyperventilation and after exercise were nearly identical. These results suggested that exercise increases the bronchoconstriction produced by a given concentration of SO2 in subjects with asthma by increasing the minute volume of ventilation and that the delay in bronchoconstriction after exercise is due to the large tidal volumes that persist for some minutes during recovery.

Numerous criteria for the diagnosis of inferior wall myocardial infarction by electrocardiogram (ECG) and vectorcardiogram (VCG) have been published, but they have not been subjected to a systematic, independent evaluation. Accordingly, we studied 146 patients undergoing cardiac catheterization; 63 were normal and 83 had a history of infarction, a significant right coronary lesion and an inferior wall motion abnormality (inferior infarction group). No ECG or VCG criteria were considered in the designation of the two groups; rather, three sets of ECG and VCG criteria were evaluated for this purpose. Specificity was excellent (98-100%) and sensitivity was poor (4-34%) by all three sets of ECG criteria, but the 1949 ECG criteria of Meyers et al. are the least sensitive (4%, p less than 0.001). Specificity (90-100%) and sensitivity (82-84%) were very good by all three VCG criteria. The VCG criteria of Starr et al. gave no false-positive results in our normal group. Because of enhanced sensitivity, the overall accuracy of the VCG was higher than that of the ECG for the diagnosis of inferior infarction (90% vs 62%, p less than 0.001). We conclude that more recent ECG criteria for the diagnosis of inferior wall myocardial infarction are highly specific, but insensitive compared with VCG criteria.

To assess the degree to which high-cost illness is terminal illness, survival data were obtained on a random sample of 776 adult high-cost patients from nine acute-care hospitals in the San Francisco Bay area. Two years after discharge, at least 34% of high-cost patients had died. These patients accounted for 39% of the hospital charges incurred by the total group of patients studied. Among the factors significantly associated with death were a diagnosis of cancer (62% death), age older than 64 years (44% death), discharge from a medical (as opposed to surgical) service (42% death), and hospital bill of $10,000 and more (48% death). In-hospital and posthospital death rates varied greatly among the nine hospitals. The data suggest that large amounts of medical care are consumed by patients with unfavorable prognoses.