Bronchoalveolar lavage (BAL) with Gomori methenamine silver (GMS) stai
n is commonly used to detect Pneumocystis carinii and fungal organisms
as causes of infectious pulmonic disease in immunosuppressed patients
. However, several reports have indicated that GMS stains are not any
more sensitive than conventional cytologic stains in detecting Pneumoc
ystis organisms in select patient populations, such as those with acqu
ired immunodeficiency syndrome (AIDS). To examine the utility of GMS s
tains in our laboratory, we retrospectively reviewed 243 BALs from 188
patients. Sensitivity of the GMS stain for Pneumocystis and for fungi
detection was 100%. Sensitivity for Pneumocystis and for fungi detect
ion by Papanicolaou stain alone was 79% and 88%, respectively; by Diff
-Quik stain alone it was 68% and 88%, respectively; and by combined Pa
panicolaou and Diff-Quik stains it was 79% and 100%, respectively. In
four additional cases, fungi were detected by other methods (culture,
biopsy) and not by BAL. The GMS stain result was correlated with a num
ber of risk variables to determine which variables were associated wit
h GMS positivity. Using stepwise logistic regression, Pneumocystis pos
itivity by GMS stain correlated (P < 0.0001) only with the variable of
history of AIDS or AIDS risk factors. Fungal organism positivity by G
MS stain correlated (P = 0.02) only with the variable of history of BA
L positivity for fungus. Cost savings analyses were performed, estimat
ing the cost of the GMS stain at $45 (total cost of GMS in 243 BALs wa
s $10,935). A maximum cost of $9,623 could have been saved by screenin
g conventional stains before using GMS stains and by using GMS stains
only in patients with the variables of history of AIDS or AIDS risk fa
ctors and history of BAL positivity for fungus.