Use of Postdischarge Emergency Medical Services to Reduce Hospital Readmissions: Does It Work and Is It Economically Feasible?
Authors: Geskey JM, Grile C, Jennings N, Good H, Crawford A, Kaminski M.
Published in: Popul Health Manag, 2019 Oct 7.
Community paramedicine is a hot topic in EMS, and community paramedics are being utilized in many ways to improve patient outcomes. This month we review a study that sought to determine whether home visits by EMS personnel could reduce unscheduled ED visits and hospital readmissions for patients who were at increased risk for hospital readmissions. The authors had a secondary objective of examining the costs associated with implementing a community paramedic program.
This retrospective case-control study was conducted in central Ohio. The study period was from January 2018 to April 2019. In a case-control study, patients who receive the study intervention (cases) are compared to patients who don’t (controls).
In this study cases were identified using the LACE index. LACE stands for length of stay, acuity of admission, comorbidities, and ED visits. LACE is a validated measure used to quantify the risk of death or unplanned readmissions within 30 days of discharge. LACE scores can range from 1–19. In this study, if a patient had a LACE score of 9 or more, their information was communicated to EMS. The patient was then scheduled for an EMS home visit within 36 hours after discharge. Those who consented were the study cases.
During the home visits, the EMS providers ensured patients understood their discharge paperwork, including follow-up appointments; performed medication reconciliation; assisted with medication organization; performed a home safety inspection; and assessed the patient’s mobility. All cases were paired with controls—in other words, every patient who received the intervention was matched with a patient similar in age, sex, insurance status, diagnosis, LACE score, and whether they had case-management interventions in addition to the EMS home visit. Matching is typically done in case-control studies to attempt to eliminate confounding variables that can distort outcomes.
The authors also evaluated total charges, estimated reimbursement, variable costs, contribution margin, ﬁxed costs, total costs, estimated proﬁts, and lengths of stay for patients who were admitted. In describing how they calculated cost, the authors noted, “EMS costs were based on salary and beneﬁts for the personnel who performed the home visit, which were converted into an hourly rate. This rate was multiplied by the amount of time spent on the intervention, after which the mileage reimbursement EMS receives (based on federal standards) was subtracted to arrive at a total cost of the intervention.”
Before we discuss the results, it should be noted this project was originally designed as a quality improvement project. There are some important differences between research and quality improvement projects. Briefly, data collection in research is more rigid, while quality improvement data collection methods can be more flexible. Often quality improvement projects inform research—this study is a great example of that.
There were 53 patients in the intervention group and 53 matched controls. When evaluating 30-day ED visits, 17% of the cases had an ED visit within 30 days, vs. 25% of the controls. While this reduction seems to suggest the intervention was successful, the difference was not statistically significant (p=0.34). Similarly, when evaluating reductions in 30-day readmissions, 19% of the cases were readmitted, vs. 26% of controls—but again the results were not statistically significant (p=0.35).
The lack of statistical significance can be explained by the limited number of patients in the analysis. If this study originated as a research project, the authors would have had to conduct a sample size calculation before collecting data to achieve sufficient power (i.e., assure they collected enough data to detect a statistically significant difference, if one existed).
Because this was a quality improvement study, that calculation was not performed beforehand. However, as we discussed above, quality improvement studies often inform research. In order to perform a sample size calculation, statisticians need to know what difference to expect between groups. This study contributed the required information to calculate a sample size, and with their results the authors determined the required sample size to evaluate the difference in ED visits was 605 patients and to evaluate readmissions 644 patients.
The evaluation of financial impact to the hospital for the control group revealed a loss of $9,915. With patients who had an EMS visit, the hospital made a $3,626 profit, while the total cost for EMS to provide the home intervention was $1,937.
This was an interesting study that will inform future community paramedicine research. Its most notable limitation is its small number of patients, which not only limits the analyses but also the generalizability of these results.
The authors also acknowledge the use of the LACE index, as opposed to other risk assessments tools, may have limited the ability to uncover differences between groups. The LACE index has been shown to have less discriminatory ability to predict hospital readmissions. The authors also note a more robust estimation of cost was possible but not performed in this study.
The results of these authors’ quality improvement project will serve as a starting point for many future investigations. When you do something impactful within your system, publish it!
Antonio R. Fernandez, PhD, NRP, FAHA, is a research scientist at ESO and an assistant professor in the department of emergency medicine at the University of North Carolina–Chapel Hill. He is on the board of advisors of the Prehospital Care Research Forum at UCLA.