Abstract
Objectives
To evaluate the quality of culture follow-up after emergency department (ED) discharge in patients with urinary tract infections (UTIs).
Methods
This convergent mixed methods study included an observational cohort study and a qualitative interview study in UTI patients discharged from the ED of a Dutch university hospital. The primary outcomes of the observational study were the proportion of patients requiring adjustment of antibiotic therapy after culture review, and the proportion of patients in whom these adjustments were made. Logistic regression identified factors associated with these outcomes. Interviews assessed patient experiences and transcripts were analysed using inductive thematic content analysis. Integration of the results informed recommendations for high-quality follow-up.
Results
Out of 455 patients, 285 (63%) required culture-based treatment adjustments. In most patients, no adjustments were made (239/285, 84%). De-escalation was most frequently omitted (98%), followed by discontinuation of antibiotics (92%). A mean of 7.1 (SD 3.8) antibiotic days per patient could have been avoided in 103 patients. Patients with diabetes were less likely to require adjustments (aOR 0.50, 95%-CI 0.29–0.85). Patients with moderate or severe renal impairment (aOR 4.1, 95%-CI 1.45–11.33; aOR 4.2, 95%-CI 1.50–11.94) or recurrent UTIs (aOR 5.0, 95%-CI 2.27–11.18) were more likely to have received necessary adjustments. Twelve interviews also revealed varying degrees of follow-up. Three themes were identified: ‘information and communication’, ‘coordination and accessibility of care’ and ‘individual needs and preferences’. Recommendations for high-quality follow-up advocate a person centred approach.
Conclusions
This study highlights the importance of urine culture follow-up after ED discharge, mainly to reduce unnecessary antibiotic treatment, promote de-escalation and improve patient experience.
Introduction
A large proportion of hospital-prescribed antibiotics are initiated at emergency departments (EDs). The predominantly empirical nature of antibiotic treatment and the need for rapid decision-making favour the use of broad-spectrum antibiotics in the ED.1 Inappropriate and unnecessary use of antibiotics is further encouraged by the fact that de-escalation and discontinuation of treatment based on culture results occurs outside the ED—patients are either admitted to the hospital or discharged home.1,2 The latter group is at risk of being overlooked by healthcare professionals, reducing the likelihood of therapy reassessment being performed.3–6 Both inappropriate prescribing in the ED and failure to reassess and if necessary adjust therapy7 make EDs a critical focus for antimicrobial stewardship programmes (ASPs). However, ASPs focusing on the ED are relatively under-developed.8
Urinary tract infections (UTIs) account for a significant proportion of infectious disease-related ED visits.9 Urine cultures are obtained in most cases of UTIs, making them pre-eminent infections that may require adjustment of initial post-discharge management. However, post-discharge follow-up is often a logistically challenging and time-consuming process. Follow-up is currently limited to pharmacist-led programmes, which have only been studied in the USA.10–16 Nonetheless, pharmacist-led stewardship intervention studies have shown that follow-up programmes for patients with UTIs discharged from the ED can improve appropriate antibiotic prescribing and reduce antibiotic overuse in patients with asymptomatic bacteriuria.10,12–16 Although these studies mainly focus on adjusting therapy after positive culture results, some studies have suggested that follow-up should also focus on discontinuing antibiotics in case of negative cultures.17–19 Culture-guided de-escalation of antibiotic therapy, although an important stewardship objective, remains underexplored in patients discharged from the ED.20,21 Finally, patient experience and engagement are increasingly recognized as a meaningful focus for ASPs.22,23 However, research in this area is limited, and in particular there is a lack of insight into how patients experience the post-ED period and what their expectations are for their follow-up.
Our aim was to evaluate the quality of follow-up of patients with suspected UTIs after discharge from the ED by assessing whether adjustments in therapy are needed based on culture results, whether these adjustments are made and what factors are associated with the need for and making of adjustments, while also shedding light on patients’ experiences in this context.
Patients and methods
Mixed methods design
This convergent mixed methods study consisted of an observational cohort study and a qualitative interview study exploring patient experiences, to illustrate and more completely understand the findings of the observational study.24 Both studies involved patients with suspected UTIs discharged from the ED of a Dutch university hospital. Interviews were conducted in parallel with the observational study. Interviewed patients were not a subsample of the records screened for the observational study. The observational study included retrospective chart review data from the period between January 2019 and April 2023. Interviews were conducted in two phases: first between October and November 2022 (pilot interviews) and later between May and July 2023 (final interviews, reported in this paper). After completing both studies, the results were integrated. We summarized for each of the interviewees whether and how their follow-up took place, illustrating their trajectory with quotes about their experiences, to ‘colour’ the findings of the observational study. As the interview findings helped to illustrate the chart review results, we also used our increased understanding to formulate recommendations for improving follow-up from the patient’s perspective.
Observational cohort study
Study population and design
All adult patients (≥18 years) who were discharged home from the ED, from whom a urine culture was obtained, and who had a positive urinalysis (positive nitrite and/or ≥20 leukocytes per high-power field) were eligible for inclusion. Eligible patients were identified using an algorithm in the electronic health record (EHR). Individual patients were included more than once only if the respective visits were more than 3 months apart. To ensure inclusion of patients with a truly suspected UTI, we used the criteria outlined in Figure S1 (available as Supplementary data at JAC Online). The standard of care in our hospital is described in Supplementary file 1.
Quantitative data collection and outcomes
Data on patient demographics, clinical characteristics and antibiotic therapy were retrospectively retrieved from the EHR. Data extraction was performed by three investigators (R.T., M.v.Z. and T.T.) using an electronic case report form (eCRF) with warnings and blocks for out-of-range values. A random sample of cases performed by M.v.Z. or T.T. was double-checked by R.T. In addition, all inconclusive cases were discussed in a team meeting (R.T., J.t.O. and J.H.) to reach consensus on the working diagnosis.
The primary outcome measures were (i) the proportion of patients in whom an adjustment of initial management after discharge was indicated based on the urine culture result, and (ii) the proportion of patients in whom this adjustment was actually made. Possible adjustments in antibiotic therapy were initiation, discontinuation, de-escalation and change of antibiotics in case of a bug–drug mismatch (see Table 1 for definitions). These outcomes were assessed by reviewing all notes from the patient’s EHR over the following 3 months; if no evidence of follow-up was found, we concluded that the adjustment was not made. Additional definitions are described in Supplementary file 3.
Table 1.
Definitions of possible adjustments in therapy
|
Secondary outcomes included patient demographics and clinical characteristics associated with an adjustment being indicated in a patient, and characteristics associated with the adjustment actually being made in a patient. Furthermore, the amount of unnecessary antibiotic days was calculated for patients whose antibiotics were incorrectly not discontinued (subtracting the time it took for culture results to become available from the total duration of antibiotics prescribed).
Statistical analysis
Descriptive statistics were performed to describe patient baseline characteristics, to calculate the primary outcomes and to compare subgroups. To find factors associated with an adjustment being indicated or being performed in a patient, first univariate logistic regression analyses were performed to select variables (if the P value was ≤0.2) for the multivariable models. Patient demographics and clinical characteristics were used in the univariate analysis; these variables were selected because they were hypothesized to have an effect on the outcomes. Second, multivariable logistic regression with backward elimination was performed. Only statistically significant (P ≤ 0.05) variables remained in the final model. A Pearson correlation matrix analysis was performed to assess multicollinearity between independent variables and, if present, the variable with the highest correlation was deleted. All analyses were performed with SPSS (v.27.0).
Interview study
Study population and design
First, seven pilot telephone interviews were conducted with adult (≥18 years) patients with all types of suspected infection discharged from the ED to test and adjust the interview guide. Eventually, the interview guide consisted of five topics, namely patient experiences with (i) follow-up, (ii) antibiotics, (iii) cultures, (iv) contact with the responsible physician and (v) the use of the patient portal to monitor their own results including whether monitoring could be the patient’s own responsibility. The last was added after the pilot interviews. Supplementary file 4 shows the interview guide.
Second, adult (≥18 years) patients who were discharged home from the ED with a suspected UTI (i.e. UTI was listed as the working diagnosis in their ED documentation) were invited for semi-structured telephone interviews. We used a convenience sampling approach and recruited patients until data saturation was reached. Patients were informed of the study either in the ED or by telephone if they were already at home. Patients were excluded if they did not speak Dutch or if they were mentally or physically unable to participate. Patients were interviewed approximately 2 weeks after discharge to allow sufficient time for any follow-up, but also to minimize recall bias.
Qualitative analysis
All interviews were audio-recorded and transcribed verbatim by an independent transcriber. After familiarization, transcripts were coded using open coding in ATLAS.ti and were then analysed using inductive thematic content analysis. The pilot interviews were coded independently by two researchers (T.T. and R.T.) and were used to develop a codebook. The codebook was then used by one researcher (R.T.) to code the final interviews. If new codes emerged, these were discussed with a second person (M.H.). The codes were grouped into themes (R.T. and M.H.) and possible discrepancies were discussed with another author. Only the final interviews with UTI patients were used for further analysis and interpretation in this study.
Ethical considerations
The Ethics Committee of the Radboudumc (CMO) decided that their formal approval was not required and that the study could be conducted according to the protocol (file number 2022-15857). Informed consent was not required for the chart review as patients could opt out (Supplementary file 1). Verbal consent was sufficient for interviews. Data were handled in accordance with Dutch privacy policies (AVG).
Results
Observational cohort study
Study population and characteristics
During the study period, 2617 patients were discharged home and had urine cultures taken. The study population consisted of 455 patients with a clinical suspicion of UTI and a positive urinalysis (Figure 1). Men (49.7%) and women (50.3%) were equally represented. The mean age was 58.8 years (standard deviation (SD) 18.4) (Table 2). Out of the total, 278 patients (61.1%) had typical UTI symptoms. Urine cultures were positive in 252 patients (55.4%), with Escherichia coli being most common (52.8%), and 372 patients (81.8%) received antibiotic therapy with a mean duration of 10.2 days (SD 4.7).
Figure 1.
Flowchart of patient inclusion observational study. Patients were initially identified based on a urine culture being performed and being discharged from the ED. Patients were then excluded if they did not have a urinalysis or if they had an opt out status (i.e. they did not agree to their data being reused). Patients with a negative urinalysis were then excluded. Finally, patients without a clinically suspected UTI or who had attended the ED more than once in the previous 3 months were excluded.
Table 2.
Demographics and clinical characteristics of the study population
| Total population | Adjustment indicated | No adjustment indicated | |
|---|---|---|---|
| N = 455 | N = 285 | N = 170 | |
| Demographics | |||
| Age, mean (SD) | 58.8 (18.4) | 59.0 (18.5) | 58.4 (18.3) |
| Male, n (%) | 226 (49.7) | 143 (50.2) | 83 (48.8) |
| Comorbidities, n (%) | |||
| Immunosuppression | 139 (30.5) | 79 (27.7) | 60 (35.5) |
| Recurrent UTIs | 85 (18.7) | 54 (18.9) | 31 (18.2) |
| Urinary tract instrumentation <1 month | 55 (12.1) | 30 (10.5) | 25 (14.7) |
| Presence of urinary tract catheter | 154 (33.8) | 92 (32.3) | 62 (36.5) |
| Diabetes mellitus | 67 (14.7) | 31 (10.9) | 36 (21.2) |
| Charlson Comorbidity Index, mean (SD) | 4.5 (3.1) | 4.4 (3.1) | 4.5 (3.0) |
| Presentation at ED, n (%) | |||
| Urinary tract symptomsa | 278 (61.1) | 178 (62.5) | 100 (58.8) |
| Only fever/malaise/abdominal pain/flank pain | 151 (33.2) | 89 (31.2) | 62 (36.5) |
| Only other symptoms | 26 (5.7) | 18 (6.3) | 8 (4.7) |
| Treating specialty, n (%) | |||
| Internal medicine | 130 (28.6) | 70 (24.6) | 60 (35.3) |
| Urology | 147 (32.3) | 95 (33.3) | 52 (30.6) |
| ED-physician | 81 (17.8) | 54 (18.9) | 27 (15.9) |
| Other | 97 (21.3) | 66 (23.2) | 31 (18.2) |
| Tissue invasion | 291 (64.0) | 189 (66.3) | 102 (60.0) |
| Received antibiotic treatment for UTI in the last month | 50 (11.0) | 29 (11.5) | 16 (10.7) |
| Received antibiotic treatment while presenting in ED | 62 (13.6) | 51 (17.9) | 11 (6.5) |
| MEWS >4 | 51 (13.9) | 36 (15.6) | 15 (11.1) |
| Diagnostics, n (%) | |||
| Positive urine cultureb | 252 (55.4) | 131 (46.0) | 121 (71.2) |
| Escherichia coli | 133 (52.8) | 73 (55.7) | 60 (49.6) |
| Klebsiella pneumoniae | 37 (14.7) | 15 (11.5) | 22 (18.2) |
| Proteus mirabilis | 13 (5.2) | 10 (7.6) | 3 (2.5) |
| Enterococcus faecalis | 11 (4.4) | 8 (6.1) | 3 (2.5) |
| Pseudomonas aeruginosa | 14 (5.6) | 9 (6.9) | 5 (4.1) |
| Staphylococcus aureus | 8 (3.1) | 4 (3.1) | 4 (3.3) |
| Candida species | 7 (2.8) | 4 (3.1) | 3 (2.5) |
| Aerococcus urinae | 7 (2.8) | 7 (5.3) | 0 |
| Other | 47 (18.7) | 18 (13.7) | 29 (23.9) |
| CRP (mg/L), mean (SD) | 57.0 (65.1) | 60.9 (71.7) | 50.6 (51.9) |
| CKD-EPI-GFR (mL/min/1.73 m2), mean (SD) | 68.3 (22.1) | 69.0 (22.1) | 67.2 (22.0) |
| Leukocytes (×1000 cells per µL), mean (SD) | 10.2 (4.3) | 10.2 (4.4) | 10.3 (4.2) |
| Blood culture performed, n (%) | 248 (54.5) | 156 (54.7) | 92 (54.1) |
| Positive, n (%) | 27/248 (10.9) | 13/156 (8.3) | 14/92 (15.2) |
| Same pathogen as found in UC, n (%) | 18/27 (66.7) | 9/13 (69.2) | 9/14 (64.3) |
| Antibiotic therapy, n (%) | |||
| Single administration of IV antibiotics at EDc | 44 (9.7) | 27 (10.7) | 15 (10.1) |
| Antibiotic treatment initiated or continued at ED | 372 (81.8) | 252 (88.4) | 120 (70.6) |
| Antibiotic treatment duration in days, mean (SD) | 10.2 (4.7) | 10.1 (4.2) | 10.3 (5.6) |
| With tissue invasion | 10.5 (4.0) | 10.6 (4.2) | 10.3 (3.7) |
| Without tissue invasion | 9.1 (6.0) | 8.7 (3.7) | 10.0 (9.3) |
BC, blood culture; CKD-EPI-GFR, Chronic Kidney Disease Epidemiology Collaboration glomerular filtration rate; CRP, C-reactive protein; MEWS, modified early warning score; UC, urine culture; SD, standard deviation.
aDysuria, haematuria, pollakiuria, increased urge, urinary retention, altered urine aspect (cloudiness, colour, smell), incontinence.
bUrine culture could be positive with multiple pathogens.
cPatients sometimes received a single dose of IV antibiotics in the ED before being discharged home.
Primary outcomes: proportion of adjustments indicated and executed
Out of 455 patients, 285 (63%) required adjustment of antibiotic therapy (Table 3). Discontinuation of treatment was most commonly indicated (55%). In most patients, the adjustments were not made (239 of 285 patients, 84%). De-escalation was most frequently omitted (98%), followed by discontinuation (92%). Forty-three (27%) of the patients whose antibiotic treatment could have been discontinued, had received antibiotic treatment before the urine culture was taken. Excluding these 43 patients from the analysis, a mean 7.1 (SD 3.8) antibiotic days per patient could have been avoided in 103 patients.
Table 3.
Primary outcome: adjustment indicated and execution of indicated adjustment based on urine culture results
| N = 455 | Number (%) |
|---|---|
| Adjustment indicated | 285 (62.6) |
| Discontinuation | 158 (55.4) |
| Initiation | 35 (12.3) |
| Bug–drug mismatch | 42 (14.7) |
| De-escalation | 50 (17.5) |
| Indicated adjustment executed | 46/285 (16.1) |
| Discontinuation | 12/158 (7.6) |
| Initiation | 10/35 (28.6) |
| Bug–drug mismatch | 23/42 (54.8) |
| De-escalation | 1/50 (2.0) |
Secondary outcomes: factors associated with an adjustment being indicated in a patient
Diabetes mellitus, immunosuppression, urinary tract instrumentation within the last month, treating physician, CRP-value and tissue invasion were included in the multivariable regression analysis (Table 4). After backward elimination, only patients with diabetes mellitus were less likely to require an adjustment in therapy [adjusted odds ratio (aOR) of 0.5 (95%-CI 0.29–0.85)]. The full analysis is shown in Table S1(a) and (b). In diabetes mellitus patients, fewer bug–drug mismatches led to less frequent adjustments in antibiotic therapy (Table S2a).
Table 4.
Secondary outcome: factors associated with an adjustment being indicated in a patient
| Univariate analysis | Multivariable analysis (n = 423) | ||||
|---|---|---|---|---|---|
| N | OR (95% CI) | P value | OR (95% CI) | P value | |
| Diabetes mellitus | 455 | 0.45 (0.27–0.77) | 0.003 | 0.50 (0.29–0.85) | 0.011 |
| Immunosuppression | 455 | 0.70 (0.47–1.06) | 0.090 | ||
| UT instrumentation | 455 | 0.68 (0.39–1.21) | 0.188 | ||
| Treating physician | 455 | 0.100 | |||
| Internal medicine (ref.) | 1 | ||||
| Urology | 1.57 (0.97–2.54) | 0.069 | |||
| ED-physician | 1.71 (0.96–3.10) | 0.067 | |||
| Other | 1.83 (1.10–3.16) | 0.032 | |||
| CRPa | 423 | 1.00 (1.00–1.05) | 0.116 | ||
| Tissue invasion | 455 | 1.31 (0.89–1.94) | 0.175 | ||
CI, confidence interval; CKD-EPI-GFR, Chronic Kidney Disease Epidemiology Collaboration glomerular filtration rate; CRP, C-reactive protein.
aCRP was not measured in all patients. Only patients with complete measurements for the multivariable model were included (n = 423).
Secondary outcomes: factors associated with an adjustment being executed in a patient
CKD-EPI-GFR (eGFR), recurrent UTIs and treating physician were included in the multivariable logistic regression analysis (Tables 5 and S1c and d). Patients with recurrent UTIs had an aOR of 5.03 (95%-CI 2.27–11.18), meaning that the odds of an adjustment being made were higher in patients with recurrent UTIs compared to patients without recurrent UTIs. Patients with moderate or severe renal impairment (eGFR 30–90 mL/min/1.73 m2) had higher odds for the adjustment being made (aOR 4.1, 95%-CI 1.45–11.33; aOR 4.2, 95%-CI 1.50–11.94). Detailed analyses of these groups showed that patients with recurrent UTIs were more likely to have all four types of adjustment made. In patients with renal impairment, execution of adjustments in therapy were most common in the eGFR 30–60 group (Table S2b and c).
Table 5.
Secondary outcome: factors associated with an adjustment being executed in a patient
| Univariate analysis | Multivariable analysis (n = 263) | ||||
|---|---|---|---|---|---|
| N | OR (95% CI) | P value | OR (95% CI) | P value | |
| Recurrent UTIs | 285 | 3.63 (1.82–7.43) | <0.001 | 5.03 (2.27–11.18) | <0.001 |
| Treating physician | 285 | 0.162 | |||
| Internal medicine (ref.) | 1 | ||||
| Urology | 1.10 (0.50–2.40) | 0.819 | |||
| ED-physician | 0.26 (0.07–0.96) | 0.043 | |||
| Other | 0.88 (0.36–2.12) | 0.771 | |||
| CKD-EPI-GFRa | 263 | 0.043 | 0.025 | ||
| >90 (ref.) | 1 | 1 | |||
| >60–<90 | 3.18 (1.19–8.50) | 0.021 | 4.05 (1.45–11.33) | 0.008 | |
| >30– ≤ 60 | 4.21 (1.55–11.44) | 0.005 | 4.23 (1.50–11.94) | 0.006 | |
| ≤30 | 2.93 (0.66–13.10) | 0.160 | 1.62 (0.34–7.81) | 0.546 | |
CI, confidence interval; CKD-EPI-GFR, Chronic Kidney Disease Epidemiology Collaboration glomerular filtration rate.
aCKD-EPI-GFR was not measured in all patients. Only patients with complete measurements for the multivariable model were included (n = 263).
Interview study
Fifteen UTI patients were invited to participate, of whom 12 (six men and six women) agreed to be interviewed (interview duration 20–45 minutes). The mean age was 55 years (range 27–71). Most patients had attended the ED several times and had experienced UTIs before. In the context of follow-up, three themes were important from the patient perspective: (i) information and communication, (ii) coordination and accessibility of care and (iii) individual needs and preferences (Table S3a and b provide an overview of coding structure and quotes).
Information and communication
Within this theme, two sub-themes were identified: the content of the information and the process by which it was conveyed. Patients described what information was important for them when being discharged from the ED, such as information about antibiotics and receiving culture results (content). They also clearly described how this information should be communicated, by describing several responsibilities for healthcare professionals (process). These responsibilities included avoiding contradictory information, taking patients seriously and being reassuring when needed.
Coordination and accessibility of care
Patients considered it important that follow-up care was both clearly coordinated and accessible. From the patient perspective, healthcare professionals should make clear agreements about follow-up and stick to them. They should take the responsibility to monitor results and contact patients timely when needed: ‘I think the doctor still needs to be more vigilant, that if there are any changes, that he/she discusses them with the patient’ (Participant-6).
Individual needs and preferences
It is important that healthcare professionals tailor information, communication and coordination of follow-up care to patients’ specific needs. Although patients generally agreed that both information and coordination were important for follow-up, individual preferences within these themes varied between patients. For example, patients had different expectations for communication. More proactive patients would ask questions themselves, while others appreciated the healthcare professional's initiative to provide information. In terms of coordination, most patients appreciated having a follow-up contact. However, some patients would initiate such a contact themselves, while others felt that having to call the hospital was a barrier: ‘[Having to call] is a barrier. But if you are called, then that barrier is immediately removed’ (Participant-5). Similarly, some patients wanted to be engaged in in agreements about their follow-up whereas others felt no responsibility to ask questions and did not want to receive explanations.
Integration of results
Table S4 provides a summary of the follow-up experienced by patients who were interviewed. Follow-up was not structured, as it was done in different ways, and sometimes not at all. It was often unclear if and how culture results would be communicated to the patient and what the consequences of these results would be. Patients had different communication preferences and expectations, that could be influenced by previous experiences. However, most patients appreciated having a (scheduled) follow-up contact. To improve follow-up from the patient's perspective, we have used our increased understanding based on the insights from both studies, to formulate recommendations for high-quality follow-up (Table 6). Supplementary file 5 provides the Good Reporting of A Mixed Methods Study (GRAMMS) checklist.
Table 6.
Recommendations for person centred follow-up after ED discharge
|
Information and communication
1. Tailor information: Tailor the information to the patient's individual needs and preferences. Some patients may need more or less information, so always consider their specific information needs. Provide information about antibiotics, cultures and follow-up. 2. Explain (follow-up of) urine culture results: Discuss that the urine culture results are pending, when they are expected to be known, and explain the possible consequences of a positive or negative culture (e.g. discontinuation, initiation, de-escalation or change in case of a bug–drug mismatch). Coordination and accessibility of care 3. Tailor follow-up to patient's individual needs and preferences: Discuss with the patient how they would like to receive results, especially if no further action is required. This may range from ‘no news is good news’ to a desire to be reassured in the event of a positive result. 4. Do not rely on tools alone: Experienced patients will often monitor their results in the patient portal, but they do appreciate explanation from a doctor and want to be kept informed of their progress. 5. Emphasize contact with hospital: Make sure that the patient clearly understands who to contact for questions, problems and follow-up. |
Discussion
This study on discharged ED patients with suspected UTIs revealed that almost two-thirds of patients required antibiotic therapy adjustment following the urine culture result. However, these adjustments were rarely made (16%). The most commonly indicated and omitted adjustment was discontinuation of antibiotics, resulting in a significant unnecessary antibiotic exposure. We also identified an opportunity to de-escalate antibiotics in 17% of patients. Patients experienced varying levels of follow-up, but almost no structured follow-up took place, although patients would have appreciated this. Taken together, this study emphasizes the importance of structured follow-up of culture results in patients discharged home from the ED with a suspected UTI.
In our study, a high number of adjustments in antibiotic therapy were indicated (62%), in contrast to many other studies where this figure ranged from 15% to 42%.12,14,25 However, this is difficult to compare because these studies often included different patient populations with multiple infection types, focused mainly on positive cultures, and reported different types of adjustments than we did. In addition, all these studies are from the USA, where the healthcare system differs from other parts of the world. A possible explanation for our high number of necessary adjustments is that we included a study population with a high suspicion of UTI and frequent empiric antibiotic prescription—and therefore a high likelihood of a positive culture—increasing the need for a treatment adjustment. Another reason for the high number of indicated adjustments may be the possibility of false-negative cultures in our discontinuation group, but even if these patients are excluded from the analysis, the proportion of indicated adjustments is still high (53%). Finally, unlike other studies, we also assessed de-escalation, which increased the number of adjustments by 17% points.
De-escalation of antibiotic treatment is an important goal of antimicrobial stewardship as it limits antibiotic selection pressure.21 A culture result should prompt a healthcare provider to evaluate and, if necessary, change treatment to a narrow-spectrum regimen.20 De-escalation remains an under-researched topic, and although this has been better studied in inpatients than in outpatients, de-escalation is often omitted there as well.26,27 Similar to our study, a study in children found that antibiotics are rarely adjusted after discharge from the ED, leading to unnecessary exposure to broad-spectrum antibiotics.28
We found that patients with diabetes mellitus were less likely to require therapy adjustment because of fewer bug–drug mismatches in this group. We expected this to be due to the presence of more resistant pathogens in diabetic patients.29 Surprisingly, this could not be confirmed in our study, and the underlying explanation remains speculative. Patients with recurrent UTIs and patients with an eGFR between 30 and 90 mL/min/1.73 m2 were more likely to have their therapy adjusted. A possible explanation is that these patients were more likely to have regular follow-up by specialists due to their medical history. We also found that patients with recurrent UTIs were more likely to have resistant pathogens (more bug–drug mismatches) and were thus more often switched to pathogen-targeted therapy.
Our interviews underscore the need for follow-up programmes. Not only is this supported by the observed random occurrence of follow-up events, but we also found that patients valued having a follow-up contact. The recommendations we made can already be applied in daily care. Given the limited understanding of the role of patients in antimicrobial stewardship, this study, along with few others, also supports patient involvement in future research.22,23 Comparing our results with the international literature on person centred care, many of the eight Picker domains are reflected in the experiences that were reported by patients.30 In particular, information provision, respect for patient preferences and care coordination appeared to be important in the follow-up context. Most patients had visited the ED before, potentially influencing their expectations, preferences and needs compared to patients visiting for the first time. Paying attention to these factors can be important to improve person centredness and can consequently lead to better patient outcomes.31
One potentially effective method of ED follow-up reported in the literature is the pharmacist-led follow-up programme, which is largely confined to the USA.10 However, we cannot simply translate this intervention to our setting, as the healthcare landscape varies significantly across continents and within countries. Future research should focus on mapping the different stakeholders responsible for follow-up in our healthcare system and seek joint solutions for the development of a follow-up programme.
This study has several strengths. First, because we only included patients with specific symptoms, a high clinical suspicion and significant pyuria, we reduced the inclusion of patients with asymptomatic bacteriuria. The inclusion of a population with a higher suspicion of UTI compared to using urinalysis alone, also resulted in a population close to the population in the RCTs in terms of culture positivity.32 Second, we addressed an important gap in literature by including de-escalation as a potential stewardship intervention for patients discharged home from the ED. Third, we included a large population compared to other studies.10 Finally, we included the patient perspective for a comprehensive assessment of the quality of follow-up. This study also has limitations. First, because of the retrospective nature of the observational study, we had to rely partly on physicians’ written reports. Possible incomplete documentation of UTI symptoms or diagnosis, which were inclusion criteria for our study, may have led to an underestimation of the true extent of the problem. Retrospective data collection may also introduce errors. We minimized this possibility by (i) using an eCRF with restrictions on entering unrealistic data, (ii) using a flowchart for inclusion, (iii) discussing cases of uncertainty, (iv) randomly checking a subsample by a second person and (v) finally checking for outliers before data analysis. Many more urine cultures were collected than the number of patients in our study. This is likely due to routine performance of urine cultures, i.e. without clinical suspicion of UTI. Second, the study was conducted in a single tertiary care centre, which may limit the generalisability of the results. Third, although the finding that over 80% of adjustments were not made is quite alarming, we lack insight into whether an adjustment might have been performed by patients’ general practitioners (GP). This is, however, unlikely, as the GP often does not receive the culture results. We did not find notes on follow-up by the GP in patients’ EHRs or in our interviews and, although we cannot draw a firm conclusion, it seems likely that there was no GP intervention. Finally, having a co-infection in addition to a UTI was not an exclusion criterium, which may have led to a slight overestimation of the number of adjustments indicated.
Conclusion
Given that nearly two-thirds of UTI patients required an adjustment in antibiotic therapy based on culture results, we emphasize the importance of a culture follow-up programme after ED discharge home. This is crucial for adjusting therapy when necessary, mainly to reduce unnecessary antibiotic treatment and promote de-escalation, but also to improve patient experience. Future research should focus on who is responsible for this follow-up and how to design and implement a structured follow-up programme in our healthcare context.
Supplementary Material
Acknowledgements
The authors would like to thank the participants from our interviews for their valuable contribution to our research.
Contributor Information
R A M Tuinte, Department of Internal Medicine, Radboud University Medical Center, Nijmegen, the Netherlands; IQ Health Science Department, Radboud University Medical Center, Nijmegen, the Netherlands; Radboudumc Community for Infectious Diseases (RCI), Radboud University Medical Center, Nijmegen, the Netherlands.
M D van Zanten, Department of Health Sciences, Faculty of Science, Vrije Universiteit Amsterdam, Amsterdam Public Health Research Institute, Amsterdam, the Netherlands.
T Takamura, Department of Internal Medicine, Radboud University Medical Center, Nijmegen, the Netherlands.
T Schoffelen, Department of Internal Medicine, Radboud University Medical Center, Nijmegen, the Netherlands; Radboudumc Community for Infectious Diseases (RCI), Radboud University Medical Center, Nijmegen, the Netherlands.
J A Schouten, IQ Health Science Department, Radboud University Medical Center, Nijmegen, the Netherlands; Radboudumc Community for Infectious Diseases (RCI), Radboud University Medical Center, Nijmegen, the Netherlands; Department of Intensive Care Medicine, Radboud University Medical Center, Nijmegen, the Netherlands.
M E J L Hulscher, IQ Health Science Department, Radboud University Medical Center, Nijmegen, the Netherlands; Radboudumc Community for Infectious Diseases (RCI), Radboud University Medical Center, Nijmegen, the Netherlands.
J J Hoogerwerf, Department of Internal Medicine, Radboud University Medical Center, Nijmegen, the Netherlands; Radboudumc Community for Infectious Diseases (RCI), Radboud University Medical Center, Nijmegen, the Netherlands.
J ten Oever, Department of Internal Medicine, Radboud University Medical Center, Nijmegen, the Netherlands; Radboudumc Community for Infectious Diseases (RCI), Radboud University Medical Center, Nijmegen, the Netherlands.
Funding
No specific funding was received for this study.
Transparency declarations
None to declare.
Supplementary data
Figure S1, Supplementary files 1 to 5 and Tables S1 to S4 are available as Supplementary data at JAC Online.
References
- 1. May L, Martín Quirós A, Ten Oever J et al. Antimicrobial stewardship in the emergency department: characteristics and evidence for effectiveness of interventions. Clin Microbiol Infect 2021; 27: 204–9. 10.1016/j.cmi.2020.10.028 [DOI] [PubMed] [Google Scholar]
- 2. Oltra Hostalet F, Núñez-Núñez M, Portillo Cano MDM et al. Analysis of quality antimicrobial agent use in the emergency department of a tertiary care hospital. Análisis de la calidad de uso de antimicrobianos en el servicio de urgencias de un hospital de tercer nivel. Emergencias 2018; 30: 297–302. [PubMed] [Google Scholar]
- 3. Vaughn VM, Hersh AL, Spivak ES. Antibiotic overuse and stewardship at hospital discharge: the reducing overuse of antibiotics at discharge home framework. Clin Infect Dis 2022; 74: 1696–702. 10.1093/cid/ciab842 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Soper NS, Appukutty AJ, Paje D et al. Antibiotic overuse after discharge from medical short-stay units. Infect Control Hosp Epidemiol 2022; 43: 1689–92. 10.1017/ice.2021.346 [DOI] [PubMed] [Google Scholar]
- 5. Vaughn VM, Gandhi TN, Chopra V et al. Antibiotic overuse after hospital discharge: a multi-hospital cohort study. Clin Infect Dis 2021; 73: e4499–506. 10.1093/cid/ciaa1372 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Callen J, Georgiou A, Li J et al. The safety implications of missed test results for hospitalised patients: a systematic review. BMJ Qual Saf 2011; 20: 194–9. 10.1136/bmjqs.2010.044339 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Denny KJ, Gartside JG, Alcorn K et al. Appropriateness of antibiotic prescribing in the emergency department. J Antimicrob Chemother 2019; 74: 515–20. 10.1093/jac/dky447 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Pulcini C. Antimicrobial stewardship in emergency departments: a neglected topic. Emerg Med J 2015; 32: 506–506. 10.1136/emermed-2014-204220 [DOI] [PubMed] [Google Scholar]
- 9. Foxman B. Epidemiology of urinary tract infections: incidence, morbidity, and economic costs. Am J Med 2002; 113(Suppl 1A): 5–13. 10.1016/S0002-9343(02)01054-9 [DOI] [PubMed] [Google Scholar]
- 10. Kooda K, Canterbury E, Bellolio F. Impact of pharmacist-led antimicrobial stewardship on appropriate antibiotic prescribing in the emergency department: a systematic review and meta-analysis. Ann Emerg Med 2022; 79: 374–87. 10.1016/j.annemergmed.2021.11.031 [DOI] [PubMed] [Google Scholar]
- 11. Goebel MC, Trautner BW, Grigoryan L. The five Ds of outpatient antibiotic stewardship for urinary tract infections. Clin Microbiol Rev 2021; 34: e0000320. 10.1128/CMR.00003-20 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Wu JY, Balmat R, Kahle ML et al. Evaluation of a health system-wide pharmacist-driven emergency department laboratory follow-up and antimicrobial management program. Am J Emerg Med 2020; 38: 2591–5. 10.1016/j.ajem.2019.12.052 [DOI] [PubMed] [Google Scholar]
- 13. Zhang X, Rowan N, Pflugeisen BM et al. Urine culture guided antibiotic interventions: a pharmacist driven antimicrobial stewardship effort in the ED. Am J Emerg Med 2017; 35: 594–8. 10.1016/j.ajem.2016.12.036 [DOI] [PubMed] [Google Scholar]
- 14. Dumkow LE, Kenney RM, MacDonald NC et al. Impact of a multidisciplinary culture follow-up program of antimicrobial therapy in the emergency department. Infect Dis Ther 2014; 3: 45–53. 10.1007/s40121-014-0026-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Trinh TD, Klinker KP. Antimicrobial stewardship in the emergency department. Infect Dis Ther 2015; 4(Suppl 1): 39–50. 10.1007/s40121-015-0084-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Santarossa M, Kilber EN, Wenzler E et al. BundlED up: a narrative review of antimicrobial stewardship initiatives and bundles in the emergency department. Pharmacy (Basel) 2019; 7: 145. 10.3390/pharmacy7040145 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Almulhim AS, Aldayyen A, Yenina K et al. Optimization of antibiotic selection in the emergency department for urine culture follow ups, a retrospective pre-post intervention study: clinical pharmacist efforts. J Pharm Policy Pract 2019; 12: 8. 10.1186/s40545-019-0168-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Woodfine JD, van Walraven C. Should emergentologists follow up on patients diagnosed with UTI having negative urine cultures? Implications for processing post-discharge laboratory results. J Eval Clin Pract 2019; 25: 260–4. 10.1111/jep.13049 [DOI] [PubMed] [Google Scholar]
- 19. Geyer AC, Draper HM, Wolf LM et al. Feasibility evaluation to expand a collaborative practice agreement and discontinue antibiotics after an emergency department or urgent care visit. Am J Health Syst Pharm 2023; 80(Suppl 4): S151–6. 10.1093/ajhp/zxad063 [DOI] [PubMed] [Google Scholar]
- 20. Abbo LM, Hooton TM. Antimicrobial stewardship and urinary tract infections. Antibiotics (Basel) 2014; 3: 174–92. 10.3390/antibiotics3020174 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. May L, Cosgrove S, L’Archeveque M et al. A call to action for antimicrobial stewardship in the emergency department: approaches and strategies. Ann Emerg Med 2013; 62: 69–77.e2. 10.1016/j.annemergmed.2012.09.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Hughes G, O'Toole E, Talento AF et al. Evaluating patient attitudes to increased patient engagement with antimicrobial stewardship: a quantitative survey. JAC Antimicrob Resist 2020; 2: dlaa046. 10.1093/jacamr/dlaa046 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Heid C, Knobloch MJ, Schulz LT et al. Use of the health belief model to study patient perceptions of antimicrobial stewardship in the acute care setting. Infect Control Hosp Epidemiol 2016; 37: 576–82. 10.1017/ice.2015.342 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Guetterman TC, Manojlovich M. Grand rounds in methodology: designing for integration in mixed methods research. BMJ Qual Saf 2024. Online ahead of print. 10.1136/bmjqs-2023-016112 [DOI] [PubMed] [Google Scholar]
- 25. Wattengel BA, Sellick JA, Mergenhagen KA. Outpatient antimicrobial stewardship: optimizing patient care via pharmacist led microbiology review. Am J Infect Control 2020; 48: 189–93. 10.1016/j.ajic.2019.07.018 [DOI] [PubMed] [Google Scholar]
- 26. De Waele JJ, Schouten J, Beovic B et al. Antimicrobial de-escalation as part of antimicrobial stewardship in intensive care: no simple answers to simple questions-a viewpoint of experts. Intensive Care Med 2020; 46: 236–44. 10.1007/s00134-019-05871-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Khasawneh FA, Karim A, Mahmood T et al. Antibiotic de-escalation in bacteremic urinary tract infections: potential opportunities and effect on outcome. Infection 2014; 42: 829–34. 10.1007/s15010-014-0639-8 [DOI] [PubMed] [Google Scholar]
- 28. Hawkins S, Ericson JE, Gavigan P. Opportunities for antibiotic reduction in pediatric patients with urinary tract infection after discharge from the emergency department. Pediatr Emerg Care 2023; 39: 184–7. 10.1097/PEC.0000000000002868 [DOI] [PubMed] [Google Scholar]
- 29. Carrillo-Larco RM, Anza-Ramírez C, Saal-Zapata G et al. Type 2 diabetes mellitus and antibiotic-resistant infections: a systematic review and meta-analysis. J Epidemiol Community Health 2022; 76: 75–84. 10.1136/jech-2020-216029 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Picker Institute Europe . The picker principles of person centred care. The eight picker principles of person centred care. 2024. https://picker.org/who-we-are/the-picker-principles-of-person-centred-care/
- 31. Doyle C, Lennox L, Bell D. A systematic review of evidence on the links between patient experience and clinical safety and effectiveness. BMJ Open 2013; 3: e001570. 10.1136/bmjopen-2012-001570 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Eckburg PB, Muir L, Critchley IA et al. Oral tebipenem pivoxil hydrobromide in complicated urinary tract infection. N Engl J Med 2022; 386: 1327–38. 10.1056/NEJMoa2105462 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.