Evidence-Based Decision Support for the Treatment of Community Acquired Pneumonia (CAP)

Scott Clark

Health Informatics, University of Adelaide

Introduction

CAP is a major cause of morbidity, mortality and health cost worldwide [ 1 2 ] . Studies of clinical practice processes and outcomes have indicated high variation in treatment processes and outcomes for CAP [ 3 4 ] .

The traditional approach to reducing treatment variation is the paper-based guideline. Guidelines generally lack the flexibility to deal with uncertain values in medicine. Taking a traditional statistical approach they divide patients into discrete categories that may not truly represent the likelihood of certain outcomes. They contain hidden utility judgements of the value of medical outcomes. Once fixed in a guideline these judgements do not allow for situational variables such as patient preference [5] . The adoption of guidelines by practitioners is highly variable and dependent on characteristics of the physician, the patient and the medical system [6] .

Once in place guidelines require constant maintenance to keep pace with clinical science, drug development and demographic variation of illness [7] . The cost in lag time and dollars of assembling review committees and dispersing updates for paper-based guidelines is large.

In contrast to paper-based guidelines, evidence-based decision support systems are flexible, easy to update, locally relevant, utility-based, patient specific [7] . They have also have been shown to improve clinician performance [8] .

I propose to design and implement evidence-based decision support systems at a teaching hospital to improve treatment and outcomes in CAP.

Major Hypothesis

Evidence based decision support systems will increase physician compliance with local clinical rules and reduce poor patient outcomes in comparison to current CAP guidelines.

Hypotheses

1. The treatment of CAP patients at a teaching hospital is not consistent with guideline recommendations.
2. The variation in CAP treatment processes at a teaching hospital are associated with poor outcomes.
3. Decision models that use influence diagrams and consider risk factors, severity signs, treatment efficacies and patient utility are able to more accurately predict outcomes of hospitalised CAP patients in comparison to current risk models.
4. The use of stakeholder and workflow analysis in the construction of decision support for the treatment of CAP will increase physician compliance with evidence based recommendations above that of a paper-based guideline.
5. Physician access to relevant current evidence concerning decision support recommendations will increase physician compliance with a computer-based guideline above that of a paper-based guideline.
6. Feedback of local key indicator performance data will improve physician compliance with decision support recommendations.

Methods

Data collection

A recent audit of CAP processes and outcomes was carried out as part of the guideline implementation and evaluation at a teaching hospital. The resulting database contains information on historical risk factors, symptoms, signs, investigation results, processes and outcomes for in excess of 650 patients. This database will be supplemented by case note review where required.

An extensive literature search and initial review of CAP processes and outcomes has already been performed. Systematic review and meta-analysis of this data will provide an evidence-base to compare current knowledge to relationships revealed by the teaching hospital data. This evidence base will be filtered and used as current best evidence attached to the computerised CAP guideline.

Hypothesis 1

I will assess the levels of compliance with process recommendations made in the teaching hospital CAP guideline using the process information in the teaching hospital CAP database.

Hypothesis 2

Using a case-control design, I will investigate the relationship between key process indicators, symptoms, signs, investigation results and outcomes in the teaching hospital CAP database. Significant predictor variables have been identified by literature review. Outcome measures include presence of complications, time to resolution of symptoms, length of stay, mortality, readmission, and costs.

Hypothesis 3

A decision model (see figure 1) considering predictor variable dependencies has already been constructed using influence diagrams [9] . This model uses probability values from literature review for risk factors, severity signs, complications, treatment efficacies and outcomes. This will be supplemented by values from the teaching hospital database. Cost is the current utility measure, however, the model will be expanded to include quasi-utility functions to represent patient preference[ 10 ] .

Figure 1: Decision model for the treatment of CAP

The stability of the predictions of this model will be assessed by sensitivity analysis [11] . The performance of this model in comparison to that of the current teaching hospital guideline, and other current prediction models such as the pneumonia severity index [12] and the BTS pneumonia prediction rule [13] will be simulated using the teaching hospital data.

Hypothesis 4

Qualitative analysis of stakeholder interview will be used for assessment of stakeholder needs and preferences in decision support design. Observational qualitative analysis will be used to assess hospital workflow around CAP patients. This data will be integrated with process timing from the teaching hospital CAP database. These sources of data will be used at various phases of intervention development to minimise development risk and maximise stakeholder buy-in as set out in the Health Implementation Process (HIP) [14] .

My initial workflow and stakeholder assessments have identified a need for an investigation result alerting system in the teaching hospital ED. I intend to implement an alerting system with local clinical rule based filtering. Further development of this intervention will follow the HIP [14] . In a pilot trial the use of this system will be assessed by touch screen activation. I will compare time to result access, time to key processes and outcomes before and after implementation.

Hypothesis 5

I have constructed a graded evidence base via literature search that will be updated daily from the PubMed database using automatic filtering software. The evidence behind computer based decision support recommendations will be accessible via mouse click on recommendation. I shall compare compliance with decision support before and after evidence access is implemented. I will also assess physician satisfaction with the evidence-base and its recommendations via questionnaire.

Hypothesis 6

I will compare the levels of compliance with the computer based decision support system before and after the implementation of feedback of prospective data on key indicator performance.

References

1. Bartlett JG, Breiman RF, Mandell LA, File TM, Jr. Community-acquired pneumonia in adults: guidelines for management. The Infectious Diseases Society of America. Clin Infect Dis 1998;26(4):811-38.

2. Guest JF, Morris A. Community-acquired pneumonia: the annual cost to the National Health Service in the UK. Eur Respir J 1997;10(7):1530-4.

3. Fine MJ, Smith MA, Carson CA, Mutha SS, Sankey SS, Weissfeld LA, et al. Prognosis and outcomes of patients with community-acquired pneumonia. A meta-analysis [see comments]. Jama 1996;275(2):134-41.

4. Fine MJ, Stone RA, Singer DE, Coley CM, Marrie TJ, Lave JR, et al. Processes and outcomes of care for patients with community-acquired pneumonia: results from the Pneumonia Patient Outcomes Research Team (PORT) cohort study. Arch Intern Med 1999;159(9):970-80.

5. Protheroe J, Fahey T, Montgomery AA, Peters TJ. The impact of patients' preferences on the treatment of atrial fibrillation: observational study of patient based decision analysis [see comments]. Bmj 2000;320(7246):1380-4.

6. Halm EA, Atlas SJ, Borowsky LH, Benzer TI, Metlay JP, Chang YC, et al. Understanding physician adherence with a pneumonia practice guideline: effects of patient, system, and physician factors. Arch Intern Med 2000;160(1):98-104.

7. Owens DK, Nease RF, Jr. A normative analytic framework for development of practice guidelines for specific clinical populations. Med Decis Making 1997;17(4):409-26.

8. Hunt DL, Haynes RB, Hanna SE, Smith K. Effects of computer-based clinical decision support systems on physician performance and patient outcomes: a systematic review. JAMA 1998;280(15):1339-46.

9. Owens DK, Shachter RD, Nease RF, Jr. Representation and Analysis of Medical Decision Problems with Influence Diagrams. Medical Decision Making 1997;17:241-262.

10. Glasziou P, Hilden J. Test selection measures. Med Decis Making 1989;9(2):133-41.

11. Clemen RT. Making Hard Decisions . Belmont, CA: Wadsworth, 1996.

12. Fine MJ, Auble TE, Yealy DM, Hanusa BH, Weissfeld LA, Singer DE, et al. A prediction rule to identify low-risk patients with community-acquired pneumonia [see comments]. N Engl J Med 1997;336(4):243-50.

13. Lim WS, Lewis S, Macfarlane JT. Severity prediction rules in community acquired pneumonia: a validation study. Thorax 2000;55(3):219-23.

14. Pradhan M. A Structured Process for Project Implementation in Health Care, unpublished.

Home > Research > CAP

© 2004 The University of Adelaide Last Modified Wednesday, 28-Jun-2006 11:32:05 CST  Health Informatics	Copyright | Privacy | Disclaimer