Duration of antibiotics through care transitions: A quality improvement initiative

Abstract

Antibiotic resistance is increasing worldwide and can be largely attributed to excess antibiotic use. At our institution, 75% of patients were prescribed excess antibiotic days and total duration of therapy was appropriate in only 24.5% of cases per the reviewers. Choice of antibiotic was appropriate in 70.4% of cases.

BACKGROUND

Antibiotic resistance is increasing worldwide, largely driven by excessive antibiotic use.^1–4 Antibiotic stewardship (AS) addresses antibiotic resistance by ensuring that patients receive the appropriate antibiotic at the correct dose, route of administration and duration. AS interventions have traditionally focused on acute care, long-term care, and ambulatory settings. However, as patients transition from one care setting to another, AS interventions should address antibiotic orders between the hospital and the home.^5,6 Improved adherence to recommended antibiotic prescribing practices prior to and at hospital discharge and through completion of therapy is an important but under-addressed target for AS. The purpose of this study was to determine the appropriateness of a total course of antibiotics to aid in prioritizing AS interventions.

MATERIALS AND METHODS

This retrospective cohort study included adult patients aged 18 years or older discharged from the Johns Hopkins Hospital between July 1, 2020 and August 31, 2020 on oral antibiotics, excluding prophylaxis for pneumocystis. All antibiotic prescribing data, including pre-admission, during admission, and at hospital discharge, as well as information on indication, was collected from the electronic medical record (EMR). All antibiotic regimens underwent adjudication by designated reviewers, who included infectious diseases physicians (SK) and pharmacists (CS, KD). Assessment of antibiotic indication, appropriateness of antibiotics, as well as treatment duration were performed. Reviewer determination of the antibiotic indication was compared to the diagnosis listed on the discharge summary. Each patient’s diagnosis was assessed for mismatch between diagnosis listed on the discharge summary and the diagnosis identified by the reviewer. A diagnosis of ‘other’ included those prescribed prophylactic antimicrobials and those with an unknown diagnosis. When unable to ascertain from chart review any clear diagnosis of infection, the diagnosis was marked as ‘unknown.’ Each antibiotic was assessed for excess duration, defined as days of therapy prescribed beyond institutional or national guideline-recommended treatment durations (ie, patient with uncomplicated community acquired pneumonia received 7 days of antibiotics would have 2 days of excess antibiotics therapy based on guideline recommended duration range of 3 to 5 days). Patients on antimicrobial prophylaxis were marked as having 0 days of excess antibiotics.

The study was reviewed by the Johns Hopkins Medicine Institutional Review Board (IRB) and deemed exempt. Descriptive statistics were used to summarize all data collected. Analyses were performed using STATA Statistical Software, version 15.

RESULTS

During the study period, 196 patients were discharged on antimicrobials. The median total duration of therapy was 14.5 days (IQR 10–24.3) Patients with durations exceeding 85 days (n=8) were all prescribed prophylactic or suppressive antimicrobials at discharge. The median inpatient antibiotic duration was 5 days (IQR 3–7), and the median post-discharge antibiotic duration was 7 days (IQR 5–14). 51 patients (26%) were prescribed antimicrobials prior to admission, including prophylaxis and suppressive therapy. Of those deemed to have excess total duration of therapy, the median duration of excess therapy was 7 days (IQR 4–13).

Each patient’s antibiotic regimen was assessed for appropriateness based on agent selected, diagnosis and duration of therapy (Table 1).

Table 1.

Appropriateness of therapy

	Appropriateness assessed from diagnosis listed on discharge summary	Appropriateness assessed from diagnosis identified by medical professional reviewer (MD/PharmD)
Entire cohort (n = 196)
Postdischarge duration, n (%)	67 (34.2)	63 (32.1)
Total duration, n (%)	53 (27.0)	48 (24.5)
Antibiotic choice, n (%)	144 (73.5)	138 (70.4)
Patients with excess antimicrobials (n = 147)
Postdischarge duration, n (%)	20 (13.6)	16 (10.9)
Total duration, n (%)	6 (4.1)	0 (0.0)
Antibiotic choice, n (%)	98 (66.7)	92 (62.6)

Antibiotic choice was appropriate in 70.4% of cases, based on antibiotic indication as ascertained by reviewers. The discharge duration for all patients was appropriate in 32.1% of cases based on antibiotic indication as ascertained by reviewers, and the total duration of therapy was identified as appropriate in just 24.5% of cases. Compared to all patients, those who received excess antibiotics at discharge had lower rates of appropriate therapy based on antibiotic indication.

Diagnoses for prescribed antimicrobials were stratified into categories (Table 2).

Table 2.

Diagnoses for patients prescribed excess duration of antibiotics

Diagnoses category	Patients with diagnosis per discharge summary, n (%)	Patients with diagnosis per reviewer, n (%)	Diagnosis mismatch, n
Lower respiratory tract infections	27 (13.8)	26 (13.3)	1
Genitourinary infections	34 (17.3)	30 (15.3)	8
Skin and soft tissue infections	44 (22.4)	46 (23.5)	8
Osteoarticular & skeletomuscular infections	15 (7.7)	12 (6.1)	3
Bloodstream infection	13 (6.6)	13 (6.6)	0
Intraabdominal/gastrointestinal infections	50 (25.5)	48 (24.5)	6
CNS infections	1 (0.5)	1 (0.5)	0
Upper respiratory tract infections	2 (1.0)	3 (1.5)	1
Prophylaxis	35 (17.9)	21 (10.7)	18
Other infections	18 (9.2)	17 (8.7)	17
Unknown	23 (11.7)	34 (17.3)	19

Excess antibiotic duration was particularly likely for an unknown diagnosis (23%), skin and soft tissue infections (16%), and antibiotic prophylaxis (12%). Diagnoses associated with diagnosis mismatch were genitourinary, skin and soft tissue, and gastrointestinal infections. Those with excess antibiotics were more likely to have a discordant diagnosis (55.1%) compared to those without excess antibiotics (12.2%).

DISCUSSION

In this study, more than half of the patients evaluated were prescribed excess antibiotics at discharge, and the total duration of antibiotics from pre-admission to post-discharge were frequently prolonged beyond guideline-recommended duration. Excess antibiotic use is a known risk factor for antimicrobial resistance, which remains a major global public health issue.^3,4 Increasing antimicrobial stewardship efforts in both the outpatient setting and at hospital or emergency department discharge can improve antibiotic prescribing by clinicians and ensure that antibiotics are only prescribed when necessary.⁶ At our institution, several interventions have been implemented to combat unnecessary antimicrobial prescribing at discharge, including patient education at discharge on antibiotics, smart sets in the EMR for antibiotics being prescribed at discharge, and education encouraging shorter courses of antibiotics where appropriate.

The results of our study are similar to those recently published assessing antibiotic prescribing patterns. Yogo et al.⁷ evaluated antibiotic prescribing patterns at time of hospital discharge and found that an estimated 53% of oral antibiotic prescriptions were inappropriate. Our results were also similar to those found by Scarpato et al.,⁸ who also evaluated antibiotic prescribing at hospital discharge and found that approximately 70% of discharge antibiotics were inappropriate in antibiotic drug choice, dose or duration. Similar to our study, both of these studies compared prescriptions with institutional or national prescribing guidelines to assess appropriateness of duration of therapy. We also included all antibiotics prescribed during our study period, while other studies excluded those receiving long-term prophylaxis as well as absence of documentation of the indication for the infection. Conner et al.⁹ and Vaughn et al.¹⁰ also evaluated antibiotic prescribing at discharge. Both of these studies identified antibiotic overuse at discharge however only evaluated common infections such as pneumonia and urinary tract infections.

Our study has several limitations. The study was conducted at a single center and included a small sample size. We used a retrospective design, and our study did lack a comparator arm. We did not assess dosing of antimicrobials, as many of the patients in our study did not have a clear diagnosis associated with antibiotic use. We used national guidelines as the standard for durations of therapy but did not account for patient-specific factors that could have influenced extending durations of therapy for certain infections.

CONCLUSIONS

In conclusion, the majority of patients at our institution were prescribed excess antimicrobials at discharge, with total durations of therapy beyond guideline-recommended durations. Understanding the total duration of antibiotic prescription, including post-discharge and pre-admission durations, is key in assessing risk from antibiotics and targeting AS interventions.