Using Readmission Rates as a Quality Indicator in Sepsis—Addressing the Problem or Adding to It? | Critical Care Medicine | JAMA Network Open | JAMA Network
[Skip to Navigation]
Sign In
Invited Commentary
Critical Care Medicine
June 4, 2021

Using Readmission Rates as a Quality Indicator in Sepsis—Addressing the Problem or Adding to It?

Author Affiliations
  • 1Division of Pulmonary/Critical Care, Rhode Island Hospital/Brown University, Providence
JAMA Netw Open. 2021;4(6):e2112873. doi:10.1001/jamanetworkopen.2021.12873

Readmissions cause a high burden to health care systems and patients, both in terms of costs to the system and patients as well as emotional burden on patients and families. The Patient Protection and Affordable Care Act (ACA) established the Centers for Medicare & Medicaid Services Hospital Readmissions Reduction Program (HRRP). Starting in 2013, a penalty was imposed as a method of providing financial incentive to hospitals for reducing preventable readmissions.1 A similar program also exists in the United Kingdom and other countries. At that time, acute myocardial infarction, congestive heart failure (CHF), chronic obstructive pulmonary disease, and pneumonia accounted for 13% of all readmissions as well as 13% of hospital costs for readmissions—a total of $7.0 billion.2 Since that time, with the increasing use of sepsis-specific International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) codes by hospital coders, sepsis is now understood as the most common cause of unplanned 30-day hospital readmissions and associated costs.3

While there are published data that indicate that using readmission rates as a quality indicator does not lead to unintended consequences or trade-offs in the quality of population health, costs of care, or the individual experience of care (the triple aim goals),4 there remains uncertainty regarding whether the readmission rate is a valid quality indicator.5 To be considered reliable, valid, and relevant, quality indicators must clearly demonstrate a beneficial effect on health and populations, with minimal influence by potential confounders, such as variability in definitions, scientific soundness, time windows, coding, and case-mix adjustment. Most importantly for using readmission rates, there must be a clear association between hospital quality of care (during and after hospitalization) and readmission rates. The published literature has reported mixed evidence that 30-day readmission rates clearly reflect hospital quality of care.4-6

Elsewhere in JAMA Network Open, Yoo et al7 published the results of their retrospective cohort study in a large population (>57 000 patients) evaluating the association between positive fluid balance in non–critically ill patients admitted to the medical floors with sepsis and 30-day readmission rates. The study analyzed data from 21 hospitals across Northern California and specifically evaluated non–critically ill patients with sepsis. The study failed to demonstrate any association between positive vs negative fluid balance and 30-day readmission rates in non–critically ill patients admitted to the medical floors. Conducting a study involving fluid balance on the medical floors is daunting at best, given the challenges of monitoring and recording accurate fluid balance in these patient areas. Therefore, it should come as no surprise that this was the major limitation of this study. In addition, the negative intake and output (I/O) patient group was older, had more severe illness, had increased comorbidities, had increased severity scores, and were more likely to have preexisting heart failure (CHF) and/or chronic kidney disease (CKD), diuretic use, and less fluid volume administered.

Because of concerns about the inaccuracy of recording fluid balance on the medical floors, the authors further evaluated I/O reported in the electronic health record in a random sample of 120 patients, verifying that the reported numbers were accurate. In any observational study, there are many confounders, all of which are acknowledged by the authors, that make the interpretation of the findings challenging. Although the authors did a nice job of severity-adjusting for age, race, body mass index, CHF, CKD, and diuretic use, the question of considerable remaining and unaccounted for confounding and bias remains. In addition, the large difference in fluid administration (1.7 L vs 4.6 L) between patients with negative and positive I/O groups raises serious concerns about confounding by indication; in this case, the small amount of fluid administered in the negative I/O group may have falsely lowered the likelihood of 30-day readmission in a group that had more severe illness, was older, and was more likely to do poorly with more fluids (ie, patients with CHF and/or CKD), therefore accounting for the decreased heart failure–related readmission rate.

Nonetheless, this was a well-done study that raises important questions. What can we learn from this study? First, the authors have addressed very important questions that need to be answered in non–critically ill patients with sepsis. Unlike previous studies in critically ill patients, is net fluid balance in patients with sepsis not associated with higher rates of 30-day readmissions? Second, the lower readmission rate in patients with CHF and/or CKD with positive I/O makes the results of the study interesting and hypothesis generating. Although clearly affected by population heterogeneity and possible confounding, the findings do raise the question of whether clinician reluctance to resuscitate patients with CHF and/or CKD and sepsis is well founded. Third, given the controversy surrounding fluid administration in patients with sepsis, is there anything modifiable in hospital management of sepsis that might affect 30-day readmission rates at all? Could this be another example that challenges the use of 30-day readmission as a quality indicator for hospital care? This is an important question that must be answered before sepsis is added to the HRRP. Without question, 30-day readmissions represent a burden to health care systems. The question this study raises is, can we do anything about it?

Back to top
Article Information

Published: June 4, 2021. doi:10.1001/jamanetworkopen.2021.12873

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Levy MM. JAMA Network Open.

Corresponding Author: Mitchell M. Levy, MD, Division of Pulmonary/Critical Care, Rhode Island Hospital/Brown University 593 Eddy St, Providence, RI 02903 (

Conflict of Interest Disclosures: None reported.

Fischer  C, Lingsma  HF, Marang-van de Mheen  PJ, Kringos  DS, Klazinga  NS, Steyerberg  EW.  Is the readmission rate a valid quality indicator? a review of the evidence.   PLoS One. 2014;9(11):e112282. doi:10.1371/journal.pone.0112282PubMedGoogle Scholar
Fingar K, Washington R. Trends in hospital readmissions for four high-volume conditions, 2009–2013: Statistical Brief #196. Published November 2015. Accessed April 28, 2021.
Mayr  FB, Talisa  VB, Balakumar  V, Chang  CH, Fine  M, Yende  S.  Proportion and cost of unplanned 30-day readmissions after sepsis compared with other medical conditions.   JAMA. 2017;317(5):530-531. doi:10.1001/jama.2016.20468PubMedGoogle ScholarCrossref
Ayabakan  S, Bardhan  I, Zheng  ZE; Triple Aim and the Hospital Readmission Reduction Program.  Triple aim and the Hospital Readmission Reduction Program.   Health Serv Res Manag Epidemiol. 2021;8:2333392821993704. doi:10.1177/2333392821993704PubMedGoogle Scholar
Prutsky  GJ, Padhya  D, Ahmed  AT,  et al.  Is unplanned PICU readmission a proper quality indicator? a systematic review and meta-analysis.   Hosp Pediatr. 2021;11(2):167-174. doi:10.1542/hpeds.2020-0192PubMedGoogle ScholarCrossref
Leppin  AL, Gionfriddo  MR, Kessler  M,  et al.  Preventing 30-day hospital readmissions: a systematic review and meta-analysis of randomized trials.   JAMA Intern Med. 2014;174(7):1095-1107. doi:10.1001/jamainternmed.2014.1608PubMedGoogle ScholarCrossref
Yoo  MS, Zhu  S, Lu  Y,  et al.  Association of positive fluid balance at discharge after sepsis management with 30-day readmission.   JAMA Netw Open. 2021;4(6):e216105. doi:10.1001/jamanetworkopen.2021.6105Google Scholar
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words