Archive for January 2009

Dr. Who?

A study in this week’s Archives of Internal Medicine by Vinny Arora and colleagues found that vanishingly few hospitalized patients could name any of their hospital doctors. Should we care?I think we should.Vinny is one of the nation’s up-and-coming researchers in the field of hospital medicine, and a good friend. In this clever study, she and colleagues at the University of Chicago interviewed over 2800 hospitalized patients over 15 months. Three-fourths were unable to name even a single doctor caring for them; of those who “could” name a doctor, the majority of names were wrong. Sobering stuff.There are multiple issues at play here. At a place like the University of Chicago Hospital, hospitalized patients on the teaching service are cared for by gaggles of residents, students, and others who are increasingly working in shifts and running for the exits because of duty hours limits. Although the limits have made the…

Pay attention to Healthcare-associated pneumonia (HCAP)

In 2005, the IDSA and ATS issued guidelines on the treatment of pneumonia in hospitalized patients, and included a new categorization of healthcare-associated pneumonia (HCAP) for patients with recent interface with medical facilities (guidelines). In this prospective cohort, researchers identified 362 patients with CAP, HAP, or HCAP, and determined rates of guideline-appropriate treatment and mortality. Patients with HCAP were much less likely to receive guideline-concordant antibiotics (27%) than patients with CAP (59%) or HAP (69%), and their mortality was much higher (18%) than patients with CAP (7%) (abstract). As a reminder, HCAP patients include: nursing home or long-term care facility residents, anyone hospitalized (2+ days) in the last 90 days, or anyone undergoing hemodialysis / wound care / chemotherapy / IV antibiotics in the last 30 days. These patients should be treated with 2 anti-pseudomonal drugs and 1 anti-MRSA drug (guidelines)

CT to diagnose endocarditis?

In this small study of 37 patients with suspected endocarditis, researchers examined the diagnostic accuracy of CT (compared to TEE or surgery), in detecting endocarditis. CT (on per valve evaluation) compared favorably to TEE (PPV 93% and NPV 98%) and surgery (PPV 96% and NPV 97%) in detecting valve abnormalities (vegetations, abscesses, perforations, fistulas, or valve dehiscence) (abstract). The authors conclude CT could be usefulfor endocarditis diagnosis after an initial TEE is negative / inconclusive, or for prosthetic valves when metallic artifacts obscure valve visualization on a TEE. Additional pre-op benefits include better anatomic mapping,  and ruling in (or out) co-existing CAD (instead of angiography).

BNP vs. symptom-guided CHF management

In this multicenter trial of 500 patients >age 60, patients were randomized to CHF medical titration based on symptoms alone (goal NYHA class <2), or symptoms + BNP (goal of <2 times upper limit of normal). There was no difference in 18 month quality of life, overall survival, or survival free hospitalizations, but the BNP-guided group did have fewer CHF hospitalizations (62% vs 72%) (only found in those <age 75) (abstract). An editorialist advocates using BNP to titrate CHF medications in patients <75 years old (editorialist). Although this was an outpatient study, this data give some credence to the common practice of using BNP measurements to gauge CHF treatment success in the inpatient setting.

Benefits of hospital IT systems

In this multi-institutional cross-sectional analysis, researchers measured physician's use of IT (CPOE, decision support, and automated notes). They found automated notes decreased in-hospital mortality (15% for every 10 point increase in use), CPOE decreased MI death by 9% and CABG death by 55%, and decision support decreased in-hospital complications by 16%. They all reduced cost. Although we cannot prove causality, this is the best evidence to date to turn IT non-believers into believers (abstract)