A Practice-Based Intervention to Improve Time-to-Antibiotic Administration in Pneumonia Suspects at Mulago Hospital, Kampala, Uganda

1
A Practice-Based Intervention toImprove
Time-to-Antibiotic Administration in Pneumonia
Suspects atMulago Hospital, Kampala, Uganda

• Luke Davis, MD
• Pulmonary Critical Care Medicine
• June 3, 2008

2
Overview

• Background
• Specific Aims
• Preliminary Studies
• Study Design
• Assessment Implementation Strategy
• Measures
• Analysis Plan
• Human Subjects
• Questions

3
Mulago Hospital, Kampala, Uganda
4
Outcomes of respiratory illness at Mulago
Hospital are poor

• In-hospital mortality among TB suspects 13
• Initial medical evaluation takes 24 hours
• 10 community acquired pneumonia (CAP) suspects
• Antibiotic treatment standardized, but not timing
  of administration

5
Antibiotic timing and CAP

• Early antibiotic administration a/w improved CAP
  outcomes1
• High-intensity education at time of
  implementation of CAP QI a/w increased adherence
  to guidelines2
• Systematic QI research uncommon in
  resource-limited settings and of unknown efficacy

1 Arch Intern Med 2004 Mar 22164(6)637-44. 2
Ann Intern Med. 2005 Dec 20143(12)881-94.
6
Specific Aims

• To determine if a structured multimodal
  educational intervention can improve
• Median time to antibiotic delivery
• Hospital length of stay
• Survival to discharge
• In patients with pneumonia at Mulago Hospital

7
Research Methods

• Study population
• Adults with cough and pulmonary infiltrates
• Medical casualty ward, Mulago Hospital
• Study design
• Prospective non-experimental single-center cohort
  study

8
Preliminary Studies

• Epidemiology of pneumonia
• 10 of all admissions, 1 or more CAP
• Focus groups
• Interests of nurses and doctors misaligned
• Pharmacists not integrated into health care team

9
Assessment strategy

• Project proposed by a senior registrar
• PRECEDE PROCEED theory for assessing
• Quality of life mortality, length of stay
• Epidemiology median time to antibiotic delivery
• Education knowledge of guidelines
• Administration support of thought-leaders

Green Kreuter, Health Program Planning, 4th
ed., NY, London McGraw-Hill, 2005.
10
(No Transcript)
11
(No Transcript)
12
(No Transcript)
13
(No Transcript)
14
(No Transcript)
15
Implementation strategy
Control Intervention

• Measurement
• Baseline serial quantitative qualitative
  outcomes
• Education
• Disseminate guidelines based on local
  international literature, vetted by local thought
  leaders
• Engagement of hospital leaders
• Head of Medicine, Chief of Clinical Services,
  Hospital Director
• Social marketing
• Reminders and clinical decision support to
  doctors nurses
• Continuous Quality Improvement
• Team of chief registrar, chief nurse, chief
  pharmacist

16
Measures

• Outcomes
• Time-to-antibiotic delivery, length of stay,
  hospital mortality
• Measured q1month x 24 months
• Predictors and covariates
• Intervention, time since implemented
• Clinical and seasonal covariates
• Time-to-antibiotic delivery, health-care worker
  qualitative ratings
• Measurement semi-blinded
• Nurse records time of arrival
• Pharmacist records time antibiotic released from
  pharmacy

17
Analysis Plan

• Interrupted time series regression
• Power calculated with month 1 data for each
  intervention
• Patients admitted with a respiratory complaint
• Effect size from ? intervention coefficient
• Process evaluation through significant covariates
• plt0.05 for interventions, time-to-antibiotic
  delivery
• Important effect sizes of other covariates for
  generating hypotheses

18
Human Subjects

• Risk to patients low, potential benefit is high
• Participation of health care workers implies
  consent to process evaluation
• Data on knowledge, attitudes, and beliefs will be
  de-identified.

19
Questions