Your browser doesn't support JavaScript. Please upgrade to a modern browser or enable JavaScript in your existing browser.
Skip Navigation U.S. Department of Health and Human Services
Agency for Healthcare Research Quality
Archive print banner

Pneumonia: New Prediction Model Proves Promising

This information is for reference purposes only. It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work. Persons with disabilities having difficulty accessing this information should contact us at: Let us know the nature of the problem, the Web address of what you want, and your contact information.

Please go to for current information.

Research Findings for Clinicians


A new clinical prediction model can help clinicians determine the most appropriate care for newly diagnosed cases of community-acquired pneumonia (CAP). The model recently was announced by a team of researchers supported through a grant from the Federal Agency for Health Care Policy and Research (AHCPR).

The model stratifies patients into risk categories based on their medical history, physical examination findings, and a limited set of laboratory and radiographic results. This is a major breakthrough, since the factors used to predict risk are clearly defined and can be readily assessed at the time of patient presentation.

The model also predicts other important medical outcomes, such as length of hospitalization, admission to an intensive care unit for respiratory failure or hemodynamic compromise, and time to usual activities.

About 600,000 of the 4 million Americans who develop CAP each year are hospitalized. Because of a lack of evidence-based admission criteria and the tendency to overestimate the risk of death, many low-risk patients who could just as safely be treated as outpatients are instead admitted for more costly inpatient care.

The investigators made projections from a prospective cohort study of 2,287 CAP patients in Pittsburgh, Boston, and Halifax, Nova Scotia. They suggest that if the model had been used, 26-31 percent of the patients who were hospitalized for care could have been treated safely as outpatients, and an additional 13-19 percent could have been hospitalized only briefly for observation. The investigators validated the model for accuracy and general applicability with data on over 50,000 CAP patients in 275 U.S. and Canadian hospitals.

During the first step of patient assessment, the patient's risk level is evaluated using factors such as age, presence of other illnesses, and abnormal physical examination findings. For patients not defined as low risk in the first step, results of laboratory tests are used to further ascertain risk of death or other adverse outcomes.

The researchers caution that clinicians may need to consider factors other than risk before deciding that a patient should have home therapy. These include patient preferences, ability to maintain oral intake, history of substance abuse, cognitive impairment, and ability to independently carry out activities of daily living. Nevertheless, preliminary evidence from this study shows that applying the prediction model in clinical practice could reduce the need for hospitalization of CAP patients without jeopardizing their health and quality of care. A firm recommendation for its clinical use will depend on future prospective trials to confirm its effectiveness and safety.

Investigators also compared medical outcomes for ambulatory and hospitalized low-risk CAP patients; assessed physician and patient decision-making processes for initial site of care and length of hospital stay; documented resource use and costs of treatment; and identified preventive health care issues. Some of their conclusions follow:

  • More expensive antimicrobial therapy and longer hospital stays may not lead to better outcomes than can be obtained with less costly antimicrobial therapy and shorter stays.
  • Many patients hospitalized for CAP—up to one-fifth of them—remain hospitalized beyond the time they reach clinical stability.
  • Most low-risk CAP patients prefer home care, but physicians generally do not ask them their preference.
  • The two areas most likely to result in major cost savings for CAP are reducing admissions of low-risk patients and decreasing length of hospital stay. Researchers also recommend studying prescribing practices for antimicrobial drugs for ways to improve their cost-effectiveness for CAP.

About the Study

The study, reported in the January 23, 1997 issue of The New England Journal of Medicine, was conducted as part of the Pneumonia Patient Outcomes Research Team (PORT), a 5-year, multi-center AHCPR-supported project directed by Wishwa N. Kapoor, M.D., M.P.H., of the University of Pittsburgh School of Medicine. PORTs are a series of studies on the quality, effectiveness, and cost-effectiveness of current therapies for treating some of the most common and costly medical conditions in the United States.

The lead author of the January 23 article in New England Journal of Medicine, Michael J. Fine, M.D., M.Sc., was supported as a Robert Wood Johnson Foundation Generalist Physician Faculty Scholar. The PORT team also included other investigators with the University of Pittsburgh School of Medicine and Graduate School of Public Health, Pittsburgh Research Institute, Harvard Medical School, and Dalhousie University.

Pneumonia Study Publications

A partial list of studies from the Pneumonia PORT, shown in reverse chronological order, follows.

  • Fine MJ, Auble TE, Yealy DM, et al. A prediction rule to identify low-risk patients with community-acquired pneumonia. The New England Journal of Medicine 1997 (January 23); 336:243-250.
  • Fine MJ, Hough LJ, Medsger AR, et al. The hospital admission decision for patients with community-acquired pneumonia: Results from the Pneumonia PORT. Archives of Internal Medicine 1997 (January 13); 157:36-44.
  • Fine MJ, Medsger AR, Stone RA, et al. The hospital discharge decision for patients with community-acquired pneumonia: Results from the Pneumonia PORT. Archives of Internal Medicine 1997 (January 13); 157:47-56.

    Coley CM, Li YH, Medsger AR, et al. Preferences for home versus hospital care among low-risk patients with community-acquired pneumonia. Archives of Internal Medicine 1996 (July 22); 156:1565-1571.

  • Minogue MF, Hough LJ, Fine MJ, et al. Patients hospitalized after initial ambulatory therapy for community-acquired pneumonia. Journal of General Internal Medicine 1996 (April); 11(supplement 1):52A.
  • Fine MJ, Smith MA, Carson CA, et al. Prognosis and outcomes of patients with community-acquired pneumonia: A meta-analysis. The Journal of the American Medical Association 1996 (January 10); 275:2.

For the full bibliography of publications related to this study see Assessment of the Variation and Outcomes of Pneumonia: Pneumonia Patient Outcomes Research Team Final Report. The report is available from the National Technical Information Service (NTIS), Springfield, VA 22161, (703) 487-4650. Ask for accession number PB97-117808. It also is available from the AHCPR Clearinghouse (AHCPR Publication No. 97-N009).

Printed copies of Pneumonia: New Prediction Model Proves Promising (AHCPR Publication No. 97-R031), as well as a fact sheet for consumers, Pneumonia: More Patients May Be Treated At Home (AHCPR Publication No. 97-R030), are available by calling the AHCPR Publications Clearinghouse at (800) 358-9295. From outside the United States, call (703) 437-2078.

Prediction Model for Identification of Patient Risk for Persons with Community-Acquired Pneumonia

The prediction model consists of an algorithm, a scoring system, and a stratification table of the risk score. Select graphic file (26 KB) or text file for the model's algorithm.

Scoring System for Prediction Model

Patient characteristicPoints assigned(1)
Demographic factors
o Males: Age (in years)
o Females: Age (in years) -10
Nursing home resident: +10
Comorbid illnesses
Neoplastic disease: +30
Liver disease: +20
Congestive heart failure: +10
Cerebrovascular disease: +10
Renal disease: +10
Physical examination findings
Altered mental status: +20
Respiratory rate 30/minute or more: +20
Systolic blood pressure <90 mmHg: +20
Temperature <35 degrees C or 40 degrees C or more: +15
Pulse 125/minute or more:+10
Laboratory findings
pH <7.35: +30
BUN >10.7 mmol/L:+20
Sodium <130 mEq/L:+20
Glucose >13.9 mmol/L:+10
Hematocrit <30 percent:+10
PO2 <60 mmHg (2): +10
Pleural effusion:+10

(1) A risk score (total point score) for a given patient is obtained
by summing the patient age in years (age minus 10 for females) and
the points for each applicable patient characteristic.
(2) Oxygen saturation <90 percent also was considered abnormal.

Stratification of Risk Score for Prediction Model

RiskRisk ClassBased on
Low II70 or fewer total points
Low III71-90 total points
ModerateIV91-130 total points
High V> 130 total points

This prediction model for prognosis in patients with community-acquired pneumonia may be used to help guide the initial decision on site of care. However, its use may not be appropriate for all patients with this illness and therefore should be applied in conjunction with physician judgment.

AHCPR Publication No. 97-R031
Current as of January 1997


The information on this page is archived and provided for reference purposes only.


AHRQ Advancing Excellence in Health Care