Prescribing Habits of Housestaff: an Important Component to Effective Antibiotic Stewardship

Background: The success of an inpatient antimicrobial stewardship program (ASP) will rely on understanding the choices made by providers when choosing empiric broad-spectrum antibiotic therapy (BST). We assessed the historical patterns of piperacillin-tazobactam (PT) usage and explored the attitudes and perceptions of medical residents (MRs) that resulted in empiric BST usage.

Methods: Prior to our ASP (7/08 - 6/09), we identified patients on PT at our 120-bed teaching hospital and completed a retrospective chart review.  PT usage was considered inappropriate in hemodynamically stable patients without HA-infections or risk factors predisposing them to CA-infections with Pseudomonas.  Based on historical usage patterns, we developed discussion topics for focus groups with MRs (PGY1-3).  We coded transcripts for qualitative analysis (Atlas TI) and identified recurrent themes.  

Results: 335 patients received ≥1 dose of PT out of 3,834 admissions.  Among the 8,086 PT doses administered, empiric PT was given in 47% (158/335); 67% (104/158) were located on medicine wards.  Length of treatment correlated with length of stay (0.41 p<0.001).  In 90% (142/158), providers had cultures at PT initiation.  Positive cultures did not alter the decision to de-escalate (65% vs. 60% p=0.49).  In 75% (118/158), empiric PT use was improper.  Misuse of PT was most common in CAP, cellulitis, and CA-UTIs.   We interviewed MRs (n=31) from 4 academic programs in DC.  Recurrent themes in BST usage included being unaware of negative consequences of BST and about what constitutes HAP or an immunocompromised patient, mistakenly believing in complete empiric coverage, overestimating consequences of using too narrow of antibiotic coverage, and desiring to protect oneself in the event of patient decompensation. 

Conclusion: The prescribing habits of physicians are primarily formed during residency.  Historical review revealed frequent, prolonged, and often inappropriate BST.  MRs were comfortable with BST, and both inflated the consequences of using more narrow regimens and misconstrued indications for empiric use.  Medical education is vital for creating an effective ASP.