Predictive value of PCO2 gap in infants.

have missed our Table 1 and page 627.x Table 1 demonstrated, in the bottom row, that disease duration in terms of months from diagnosis was not different among groups. On page 627, we note in the second paragraph that similar proportions of patients with radiographic types 1, 2, and 3 received treatment with cortico¬ steroids, although fewer patients with type 0 radiographs had symptoms that required treatment. We agree with Dr. Reich that our use of the term relapse may be unusual or nonstandard when applied to the spontaneous remission group. We tried to use a single term to denote the appearance of symptoms severe enough to warrant treatment, following a sustained period without such symptoms. Indeed, we could have termed this "progression" in the spontaneous group rather than "relapse." We do agree that recurrence or relapse following complete spontaneous clinical and radiographic resolu¬ tion is an unusual event. In fact, this represents one of the main points we were trying to make. Unfortunately, no other data were available from the 1973 study cited by Dr. Reich. Dr. Reich asks what we would advocate for those patients who are either symptomatic or who are asymptomatic but who demonstrate clear evidence of radiographic or physiologic pro¬ gression. Our approach is to treat the symptomatic patients with low doses, ie, 15 to 20 mg/day, of prednisone for 6 to 12 months. Those who are asymptomatic but who demonstrate radiographic progression alone we do not treat routinely. For those patients who demonstrate clinically significant physiologic progression by pulmonary function studies, we do advocate a trial of corticoste¬ roids at the above dosage. sarcoidosis: the relationship of relapse to corticosteroid ther¬ We were interested by the study of Duke and colleagues1 of the difference between tonometer saline and arterial carbon dioxide tension (Dco2) as a predictor of outcome. We have recently studied 62 critically ill adults to assess the usefulness of the gastric intramucosal to arterial carbon dioxide gradient (Pico2-Paco2), among other parameters derived from gastric tonometry, as a predictor of outcome. Patients were entered into the study within 6 h of admission to our ICU and measurements were taken at 0, 12, and 24 h. At no time was there a significant difference in the Pico2-PaC02 between those who survived and those who died. The study had a power of 90% to detect a 1 standard deviation difference between the two groups. The …