I came across a couple of interesting retrospective studies while trying to find some evidence to help guide the decision-making process.
The first study analyzed 116 cases of pneumocystis pneumonia in non-HIV infected patients. The major common exposure was steroid treatment in the month preceding presentation (90.5% of patients) – the median dose was equivalent to 30mg prednisone for a median of 12 weeks before presentation.
The second reported risk factors in 41 cases in a French centre. Mortality was 29%; 85% had received steroids at a median daily steroid dose equivalent to 15mg prednisone. Only 17% had received prophylaxis.
The third used a matched case-control type analysis in patients with SLE – they reported 15 cases of pneumocystis infection; risk factors for infection compared to 60 matched controls included higher disease activity, renal involvement and higher mean cumulative steroid dose (49 vs 20mg/day; p