Journal of General Internal Medicine

, Volume 33, Issue 4, pp 493–493 | Cite as

Capsule Commentary on Sanz et al., A Composite of Functional Status and Pneumonia Severity Index Improves the Prediction of Pneumonia Mortality in Older Patients

  • Barbara Jones
Capsule Commentary

Severity assessment in pneumonia is a challenging activity that requires clinicians to pursue, recognize, and incorporate multiple pieces of data to make critical decisions, including resuscitation, site-of-care, and antibiotic selection. At best, bedside severity assessment tools such as the pneumonia severity index1 can aid overloaded clinicians by reducing biases and raising awareness of important contributors to severity that may be overlooked. At their worst, however, they oversimplify severity assessment, and over-reliance on them can reduce our attention to other important information.

In this issue of the Journal of General Internal Medicine, Sanz et al. report a strong relationship between baseline functional status (as measured by the Barthel Index for Activities of Daily Living) and 30-day mortality, beyond those risk factors contributing to the pneumonia severity index (PSI) among hospitalized pneumonia patients ≥65 years old.2 The authors demonstrate that a combined model incorporating functional status and PSI provides more accurate prognostication than PSI alone.

An important limitation is that their proposed model was not tested in clinical practice for effectiveness. The feasibility of incorporating the Barthel Index into clinical workflow is an important question, as calculating the PSI alone requires significant time. However, the study highlights the importance of paying attention to functional status, which is not explicitly incorporated into any of the previously validated severity assessment tools. While many clinicians likely incorporate this characteristic implicitly, providing bedside decision support that raises awareness of functional status could represent an opportunity to improve decision-making.

The real test of any severity assessment tool is not whether it accurately predicts a feared outcome, but whether it improves our decisions. As was recently advised, “We need tools that tell us what is not obvious on our assessment.”3 As we achieve greater modernization of our healthcare systems, prognostication tools will continue to improve, with promising accuracy. However, appropriate severity assessment requires a level of attention and situational awareness that cannot be achieved by using algorithms alone. Sanz’ study serves as a reminder that we must continue to seek features that might be overlooked, either by a clinician, or by our predictive tools.


Compliance with Ethical Standards

Conflict of Interest

The author has no conflict of interest with this article.


  1. 1.
    Fine MJ, Auble TE, Yealy DM, et al. A prediction rule to identify low-risk patients with community-acquired pneumonia. N Engl J Med. 1997;336(4):243–50. Scholar
  2. 2.
    Sanz F, Morales-Suárez-Varela M, Fernández E, Force L, Pérez-Lozano MJ, Martín V, Egurrola M, Castilla J, Astray J, Toledo D, Domínguez, A. A composite of functional status and pneumonia severity index improves the prediction of pneumonia mortality in older patients. J Gen Intern Med.
  3. 3.
    Waterer G. Severity scores and community-acquired pneumonia. time to move forward. Am J Respir Crit Care Med. 2017;196(10):1236–8. Scholar

Copyright information

© Society of General Internal Medicine (outside the USA) 2018

Authors and Affiliations

  1. 1.Division of Pulmonary & Critical Care Medicine University of UtahSalt Lake CityUSA
  2. 2.IDEAS Center of InnovationVeterans Affairs Salt Lake City Healthcare SystemSalt Lake CityUSA

Personalised recommendations