Lag times between onset of illness, specimen collection, and reporting dates were stable throughout the data set, except for the single batch-reporting artifact in 1995 (for which adjustment was made in the regression analysis).
Dr. Julie L. Gerberding, the C.D.C. director, said her agency worried about the precise time in the course of a patient's illness the specimens were obtained.
All four patients had detectable SARS-CoV antibodies by one or more laboratories very early in the illness; serum specimens collected 6 days after onset from patients 1 and 2 were positive by all laboratories by one or more methods, and specimens collected at 8 days from patients 3 and 4 were positive by EIA performed at laboratories A and B, respectively.
The timing of specimen acquisition may also be relevant, since the average delay between onset of illness and specimen collection was 9 days for our PCR-positive cases versus 16 days for all cases studied with broad-range PCR.
Intervals between date of illness onset, specimen date, and reporting date were reviewed to assess consistency of reporting over time.
Although rapid diagnostic tests can help confirm an influenza diagnosis, their sensitivity ranges from 40 70% depending on the day of illness and specimen type [ 15].
However, unless systematic differences were present by site in terms of the severity of illness affecting specimen collection, this variability should have been nondifferential and biased the findings toward the null.
Only one serum specimen was available for two of the patients at day 10 and day 36 of their illness; paired serum specimens were available for the other four.
Of the 7 employees of company B, none reported illness, and no specimens were tested for histoplasmosis.
All nine persons with probable cases were hospitalized and died as a result of their illness before laboratory specimens could be collected (case-fatality rate = 69%).
