Impact of an evidence-based algorithm on quality of care in pediatric parapneumonic effusion and empyema

Pediatr Pulmonol. 2011 Jul;46(7):722-8. doi: 10.1002/ppul.21429. Epub 2011 Feb 15.


Objective: To determine whether implementation of a collaborative, evidence-based algorithm for care of pediatric parapneumonic effusion and empyema (PPE) can improve the quality of care delivered.

Study design: Prospective cohort with retrospective control comparison of children aged 1 month to 18 years admitted with a clinical diagnosis of PPE. Quality improvement techniques were used to develop an algorithm, which was implemented September 2008. Primary outcome measures were decreased median and variability in length of stay (LOS), reduction in the use of chest computed tomography (CT), reduction in the total number of painful procedures, and increased initial use of effective drainage procedures when drainage was indicated.

Results: Compared with controls, algorithm implementation substantially reduced use of chest CT (0% vs. 41% of patients, P = 0.01) with no observed negative impact on LOS. Reductions in median LOS were not significant, but variability in LOS was reduced (P < 0.01 by F-test). Changes in number of procedures and use of effective drainage when indicated were in the predicted direction but not statistically significant.

Conclusions: Quality improvement techniques are an effective means for incorporating evidence-based medicine into pediatric care. PPE can be managed safely without the use of chest CT.

MeSH terms

  • Adolescent
  • Algorithms*
  • Case-Control Studies
  • Child
  • Child, Preschool
  • Cohort Studies
  • Drainage, Postural
  • Empyema, Pleural / microbiology
  • Empyema, Pleural / therapy*
  • Evidence-Based Medicine
  • Female
  • Humans
  • Infant
  • Male
  • Pleural Effusion / diagnostic imaging
  • Pleural Effusion / microbiology
  • Pleural Effusion / therapy*
  • Prospective Studies
  • Quality of Health Care*
  • Retrospective Studies
  • Tomography, X-Ray Computed / statistics & numerical data