Abstract
Background
Acute diarrhea is highly prevalent in Thailand, and inappropriate antibiotic use remains common. The effectiveness of a clinical pathway pop-up integrated into a computerized prescription system for antibiotic stewardship has not been previously evaluated.
Materials and Methods
We conducted a quasi-experimental study following implementation of a clinical pathway pop-up on 21 December 2021. Outpatient records from Srinagarind Hospital and two nearby primary care units were reviewed from November 2021 to January 2022. Patients were divided into pre- and post-intervention groups. Factors associated with antibiotic prescribing and appropriate antibiotic management were analyzed.
Results
A total of 460 patients were included (230 per group); 26.7% were male, with a mean age of 34.8, and 82% of patients were treated in the emergency department. Antibiotics were prescribed to 30.9% of patients in both pre-intervention and post-intervention groups (P>0.999). Rational antibiotic use was 76.5% before and 75.7% after the intervention (P=0.827). Segmented linear regression analysis demonstrated no significant change in prescription rates or appropriate antibiotic management before and after the intervention. Care provided by a resident physician was independently associated with lower odds of appropriate antibiotic management (adjusted odds ratio, 0.38; 95% confidence interval, 0.15–0.85; P=0.020).
Conclusion
The implementation of a clinical pathway pop-up window in a computerized prescription system did not significantly reduce inappropriate antibiotic use or improve rational prescribing in patients with acute diarrhea. Resident physicians as primary care providers were independently associated with inappropriate antibiotic management.
Trial Registration
Thai Clinical Trials Registry: TCTR20210420002
Keywords: Acute diarrhea, Antibiotic stewardship, Antibiotic prescription, Clinical pathway, Pop-up window
Graphical Abstract
Introduction
Multidrug-resistant (MDR) bacteria represent a critical and escalating global health challenge. In the United States, MDR infections were associated with approximately 99,000 deaths annually in 2016, including more than 50,000 deaths from hospital-acquired pneumonia and sepsis. These infections contributed to excess healthcare expenditures exceeding USD 8 billion per year and prolonged hospitalizations of more than 13 days [1]. In Thailand, the annual incidence of MDR infections exceeds 100,000 cases, resulting in more than 30,000 deaths and imposing a substantial economic burden. In-hospital antimicrobial-resistant infections accounted for over THB 66.4 billion (approximately 0.4% of gross domestic product) during 2015–2019 [2,3]. Inappropriate antibiotic use remains a major driver of the emergence and dissemination of antimicrobial resistance [4].
Acute diarrhea remains a high-incidence condition. According to the Division of Epidemiology, Department of Disease Control, Thailand, 685,223 cases of acute diarrhea were reported in 2023. Among these, only 2,442 patients (0.4%) were diagnosed with dysentery requiring antibiotic therapy [5]. Despite the low proportion of cases necessitating antibiotics, inappropriate antibiotic use remains common. The prevalence of antibiotic prescribing for acute diarrhea was reported to be 86.8% in Ethiopia [6], 43.0% in New Delhi, India [7], and 45.1% at King Chulalongkorn Memorial Hospital, Thailand [8]. These findings indicate substantial overuse of antibiotics for acute diarrhea worldwide, including in Thailand. In response, the Ministry of Public Health of Thailand launched a national service plan in 2016 to promote rational drug use in health-promoting hospitals and primary care units (PCU), setting a target to reduce antibiotic prescribing for acute diarrhea to below 20% [9].
Several antibiotic stewardship interventions for acute diarrhea, including training workshops, clinical practice guidelines, and educational materials, have been shown to reduce antibiotic prescriptions by 5%–40% [10–13]. Integrating a clinical pathway pop-up window into computerized prescribing systems is a novel strategy that fits well within electronic clinical workflows. A previous study at Srinagarind Hospital, Khon Kaen University demonstrated that a stewardship pop-up intervention for upper respiratory tract infections (URI) reduced antibiotic prescription rates from 53.8% to 32.0% after six weeks [14]. Antimicrobial stewardship programs have also been associated with reductions in MDR organisms, antibiotic-related adverse events such as Clostridioides difficile infection, and healthcare costs, without increasing mortality or treatment failure [15]. However, no previous study has specifically evaluated a clinical pathway–based stewardship intervention for acute diarrhea. Therefore, this study aimed to assess the impact of implementing a clinical pathway pop-up window on antibiotic prescribing for acute diarrhea and to identify factors associated with antibiotic prescribing and appropriate antibiotic management.
Materials and Methods
1. Participants and study design
We conducted a quasi-experimental study at Srinagarind Hospital, Faculty of Medicine, Khon Kaen University, Thailand, and at two affiliated PCUs: PCU 123 (Student PCU) and PCU Samliam (Community PCU), between November 2021 and January 2022. Patients aged ≥18 years with a principal diagnosis of acute diarrhea, as defined according to the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) codes, were screened and enrolled during the study period, irrespective of the indication recorded in the prescribing system. A clinical pathway pop-up window was integrated into the hospital’s computerized prescribing system on 21 December 2021, dividing the study period into two phases: a pre-intervention period (1 November 2021 to 20 December 2021) and a post-intervention period (21 December 2021 to 30 January 2022). The date of implementation was used as the index point, and data were collected both retrospectively and prospectively until the calculated sample size was achieved. When physicians attempted to prescribe norfloxacin or ciprofloxacin—two of the most commonly used antibiotics for acute diarrhea—a pop-up window appeared, prompting confirmation of whether the prescription was intended for acute diarrhea or for another bacterial infection. If acute diarrhea was selected, a second window was displayed informing prescribers that “more than 95% of antibiotic prescriptions for acute diarrhea are unnecessary.” Physicians were then required to select at least one clinical indication for antibiotic therapy, in accordance with the clinical practice guidelines for acute diarrhea issued by the Department of Disease Control, Ministry of Public Health of Thailand, and the 2017 Infectious Diseases Society of America (IDSA) clinical practice guidelines for the diagnosis and management of infectious diarrhea [16,17], by completing designated checkbox fields before proceeding to the final screen. The final screen was the standard prescribing interface, where medication dose and duration were specified. If the prescription was intended for a diagnosis other than acute diarrhea, the system bypassed the second pop-up window and directed physicians to the standard prescribing interface.
Clinical data were extracted from outpatient medical records. Patient inclusion was based on diagnostic coding, irrespective of the indication selected for antibiotic prescribing in the computerized system. Patients with a documented fluoroquinolone allergy, incorrect ICD-10-CM coding, or incomplete medical records were excluded. The collected variables included baseline demographic characteristics, clinical manifestations, visit location, physician age and specialty, as well as the type and duration of prescribed antibiotics.
The primary outcome was the comparison of antibiotic prescribing rates between the pre-intervention and post-intervention periods. Secondary outcomes included the increase in rational antibiotic prescribing following implementation of the clinical pathway pop-up window and the identification of factors associated with antibiotic prescription and appropriate antibiotic management. Additionally, an analysis at 3 months post-implementation was performed to evaluate outcomes after physicians had become familiar with the system.
2. Ethics statement
Ethical approval was provided by the Center for Ethics in Human Research, Khon Kaen University, in accordance with the Declaration of Helsinki (Number HE631319).
3. Sample size calculation
The sample size was calculated based on the primary outcome using the formula for comparison of two independent proportions. Data from a pilot study at Srinagarind Hospital demonstrated an antibiotic prescription rate of 30.9% for acute diarrhea prior to implementation of the clinical pathway pop-up window intervention. A previous study conducted at Siriraj Hospital reported a post-intervention antibiotic prescription rate of 19.1% following multiple rational drug use interventions [11]. Assuming that the intervention in the present study would reduce the antibiotic prescription rate to below 20%, in accordance with the Ministry of Public Health service plan policy, we estimated that a minimum of 230 participants per group would be required to achieve statistical significance.
4. Operational definitions
• Acute diarrhea
Defined as the presence of an ICD-10 Clinical Modification (ICD-10-CM) code ranging from A00 to A09, recorded as either the primary diagnosis or a comorbidity by the treating physician during the outpatient department (OPD) visit.
• Rational antibiotic prescription
Determined through retrospective chart review based on predefined criteria in accordance with the clinical practice guideline for acute diarrhea in adult issued by the Department of Disease Control, Ministry of Public Health of Thailand, and the 2017 Clinical Practice Guidelines for the Diagnosis and Management of Infectious Diarrhea by the IDSA [16,17]. Antibiotic use was considered rational when at least one of the following criteria was present:
• Persistent symptoms lasting more than 7–10 days without clinical improvement.
• Severe symptoms, defined as body temperature ≥38°C.
• Presence of mucoid or bloody diarrhea.
• Detection of white blood cells and red blood cells on stool examination.
5. Statistical analysis
Categorical variables are presented as frequencies and percentages, whereas continuous variables are expressed as mean±standard deviation. Differences in overall and rational antibiotic prescribing rates between the pre- and post-intervention periods were analyzed using the Pearson chi-square test or Fisher’s exact test, as appropriate. Interrupted time-series analysis using segmented linear regression was additionally performed to evaluate secular trends and intervention effects over time, including the baseline pre-intervention trend, immediate post-intervention level change, and change in post-intervention trend.
Associations are reported as odds ratios (ORs) with 95% confidence intervals (CIs). Factors associated with antibiotic prescribing and appropriate antibiotic management were initially assessed using univariate analysis. Variables with P<0.05 were subsequently included in multivariable logistic regression models to estimate crude and adjusted ORs. All statistical analyses were conducted using SPSS for Windows (version 28.0, IBM Corp., Armonk, NY, USA).
Results
Data from 460 OPD visits with a principal diagnosis of acute diarrhea were collected between 1 November 2021 and 30 January 2022. Participants were divided into pre-intervention and post-intervention groups, each comprising 230 visits. The clinical pathway pop-up window was integrated into the computer-based prescription system on 21 December 2021. The mean age of participants was 34.8 years, and 26.7% were male. Overall, 30.4% had underlying medical conditions, including hypertension (10%), dyslipidemia (5.7%), diabetes mellitus (5.2%), heart disease (5.0%), malignancy (2.6%), and psychiatric disorders (2.0%). A higher prevalence of underlying medical conditions was observed in the post-intervention group (34.8% vs. 26.1%, P=0.043).
The most common clinical presentations were nausea or vomiting (80.3%), abdominal pain or tenderness (71.3%), and dehydration (61.2%). These symptoms were similarly distributed between groups, as were the duration of illness and the frequency of diarrhea. The mean duration of illness was 20.2±26.5 hours (P=0.793), and the mean frequency of diarrhea was 6.9±5.4 episodes (P=0.203). Most patients were treated in the emergency department (ED) (82.0%), followed by the student PCU (15.0%). The majority of care providers were general practitioners (GPs) (71.7%) and residents (22.2%), with a mean physician age of 27.9 years. These characteristics were comparable between the two groups (Table 1).
Table 1. Baseline characteristic of patients between pre-intervention and post-intervention groups.
| Characteristic | Total (N=460) | Pre-intervention (N=230) | Post-intervention (N=230) | P-valuea | |
|---|---|---|---|---|---|
| Age, years, mean±SD | 34.8±17.9 | 34.8±18.3 | 34.7±17.5 | 0.649 | |
| Sex, male, n (%) | 123 (26.7) | 60 (26.1) | 63 (27.4) | 0.752 | |
| BMI, kg/m2, mean±SD | 23.3±4.5 | 23.9±4.1 | 22.8±4.8 | 0.210 | |
| Medical conditions, n (%) | 140 (30.4) | 60 (26.1) | 80 (34.8) | 0.043 | |
| Hypertension | 46 (10.0) | 26 (11.3) | 20 (8.7) | 0.351 | |
| Dyslipidemia | 26 (5.7) | 8 (3.5) | 18 (7.8) | 0.043 | |
| Diabetes mellitus | 24 (5.2) | 9 (3.9) | 15 (6.5) | 0.208 | |
| Heart disease | 23 (5.0) | 8 (3.5) | 15 (6.5) | 0.123 | |
| Malignancy | 12 (2.6) | 7 (3.0) | 5 (2.2) | 0.599 | |
| Psychiatric disorders | 9 (2.0) | 5 (2.2) | 4 (1.7) | >0.999 | |
| Systemic corticosteroids or immunosuppressant(s), n (%) | 5 (1.1) | 1 (0.4) | 4 (1.7) | 0.372 | |
| Clinical presentation, n (%) | |||||
| Nausea or vomiting | 354 (80.3) | 177 (80.1) | 177 (80.5) | 0.923 | |
| Abdominal pain/tenderness | 328 (71.3) | 158 (68.7) | 170 (73.9) | 0.216 | |
| Signs of dehydration | 120 (61.2) | 65 (67.5) | 55 (55.0) | 0.068 | |
| Fever | 28 (6.1) | 16 (7.0) | 12 (5.2) | 0.435 | |
| Mucous or bloody diarrhea | 13 (3.8) | 9 (5.4) | 4 (2.3) | 0.133 | |
| Diarrheal frequency, mean±SD | 6.9±5.4 | 6.5±4.9 | 7.2±5.9 | 0.203 | |
| Duration of illness (hours), mean±SD | 20.2±26.5 | 20.3±27.3 | 20.1±25.7 | 0.793 | |
| Revisiting case, n (%) | 31 (6.7) | 21 (9.1) | 10 (4.3) | 0.041 | |
| Diagnosis by ICD-10 code, n (%) | 0.944 | ||||
| A09 | 397 (86.3) | 199 (86.5) | 198 (86.1) | ||
| A05 | 59 (12.8) | 30 (13.0) | 29 (12.6) | ||
| A08 | 4 (0.9) | 1 (0.4) | 3 (1.3) | ||
| Site of visit, n (%) | 0.083 | ||||
| Emergency department | 377 (82.0) | 198 (86.1) | 179 (77.8) | ||
| Student PCU | 69 (15.0) | 29 (12.6) | 40 (17.4) | ||
| GP OPD | 6 (1.3) | 1 (0.4) | 5 (2.2) | ||
| Medicine OPD | 3 (0.7) | 2 (0.9) | 1 (0.4) | ||
| Premium clinic | 2 (0.4) | 0 | 2 (0.9) | ||
| Others | 3 (0.7) | 0 | 3 (1.3) | ||
| Type of physician, n (%) | 0.451 | ||||
| GP | 330 (71.7) | 163 (70.9) | 167 (72.6) | ||
| Resident | 102 (22.2) | 53 (23.0) | 49 (21.3) | ||
| Other specialists | 14 (3.0) | 9 (3.9) | 5 (2.2) | ||
| Internist | 12 (2.6) | 5 (2.2) | 7 (3.0) | ||
| Surgeon | 2 (0.4) | 0 | 2 (0.9) | ||
| Physician age, years, mean±SD | 27.9±8.2 | 28.2±8.5 | 27.6±7.8 | 0.128 | |
aP<0.05 indicates statistical significance. Clinical-presentation percentages were calculated after excluding missing data, as in the submitted manuscript. Others included subspecialty clinic and community PCU.
SD, standard deviation; BMI, body mass index; A09, other gastroenteritis and colitis of infectious and unspecified origin; A05, other bacterial foodborne intoxications, not elsewhere classified; A08, viral and other specified intestinal infections; GP, general practitioner; OPD, outpatient department; PCU, primary care unit.
1. Antibiotic prescription rate and its rationale
The overall antibiotic prescribing rate was identical between the pre-intervention and post-intervention groups (71/230 [30.9%] vs. 71/230 [30.9%]; P>0.999). Segmented regression analysis showed no significant baseline trend prior to the intervention (β=−0.007, P=0.624). The intervention was associated with a non-significant immediate increase of 5.3% in antibiotic prescribing (β=0.053, P=0.518), with no significant change in the post-intervention trend (β=−0.006, P=0.807), resulting in a post-intervention slope of −0.013 per time unit (Fig. 1).
Figure 1. Segmented regression of antibiotic prescription and appropriate management before and after intervention.
Similarly, the proportion of cases with appropriate antibiotic management did not differ significantly between the pre-intervention and post-intervention groups (176/230 [76.5%] vs. 174/230 [75.7%]; P=0.827). Segmented linear regression analysis demonstrated no significant immediate change in appropriate antibiotic management following the intervention (β=−0.011, P=0.888) and no significant pre-intervention temporal trend (β=−0.010, P=0.425). Although the post-intervention trend showed a numerical increase, this change did not reach statistical significance (β=0.031, P=0.173) (Fig. 1). After excluding patients who received antibiotics other than norfloxacin or ciprofloxacin, which were unrelated to the pop-up intervention, antibiotic prescribing rates remained similar between the two groups (66/225 [29.3%] vs. 64/223 [28.7%], P=0.883).
Among patients who received antibiotics, the proportion of rational prescriptions did not differ significantly between the pre-intervention and post-intervention groups (18/71 [25.4%] vs. 16/71 [22.5%]; P=0.694). The mean duration of antibiotic use was also comparable between the two groups. In the subgroup analysis of patients presenting to the ED, antibiotic prescription rates were similar between the pre-intervention and post-intervention groups (69/198 [34.8%] vs. 58/179 [32.4%], P=0.616), as were the rates of appropriate antibiotic management (146/198 [73.7%] vs. 132/179 [73.7%], P=0.999). Evaluation of physicians’ responses to the pop-up system during the post-intervention period showed that 29.1% of antibiotic prescriptions bypassed the pop-up without a documented indication, while 53.8% contained indications that were inconsistent with the corresponding medical records.
The most commonly prescribed antibiotic was norfloxacin (52 patients in each group, 73.2%), with similar use in both groups. The second most commonly prescribed regimen differed between groups. In the pre-intervention group, intravenous ceftriaxone combined with norfloxacin accounted for 15.5% of prescriptions, whereas ciprofloxacin monotherapy was the second most common regimen in the post-intervention group (12.7%) (P=0.048) (Table 2).
Table 2. Antibiotic prescription and its rationale on baseline and intervention period.
| Outcome | Pre-intervention | Post-intervention | P-value | |
|---|---|---|---|---|
| All cases | 230 | 230 | ||
| Antibiotic prescription, n (%) | 71 (30.9) | 71 (30.9) | >0.999 | |
| Appropriate antibiotic management, n (%) | 176 (76.5) | 174 (75.7) | 0.827 | |
| Antibiotic-prescription cases | 71 | 71 | ||
| Duration of therapy (days), mean±SD | 4.8±1.7 | 4.5±1.4 | 0.870 | |
| Rational antibiotic prescription, n (%) | 18 (25.4) | 16 (22.5) | 0.694 | |
| Type of antibiotic, n (%) | 0.048 | |||
| Norfloxacin | 52 (73.2) | 52 (73.2) | ||
| Ciprofloxacin | 3 (4.2) | 9 (12.7) | ||
| Ceftriaxone stat then norfloxacin | 11 (15.5) | 3 (4.2) | ||
| Othersa | 5 (7.0) | 7 (9.9) | ||
aOther antibiotics included cefixime, co-amoxiclav, and metronidazole.
SD, standard deviation.
At 3 months post-implementation, antibiotic prescribing rates and appropriateness were reassessed. Among 100 patients with acute diarrhea (22 March–12 April 2022), most presented to the ED (74.0%). Antibiotic prescribing and appropriate management rates were 32% and 72%, respectively, with no significant differences compared with the pre-intervention and post-intervention periods (P=0.976 and 0.675, respectively). Among patients who received antibiotics, 5 of 32 cases (15.6%) were considered appropriate, without significant differences compared with prior periods (P=0.548).
2. Factors associated with appropriate antibiotic management
In the univariate analysis, several clinical presentations were significantly associated with appropriate antibiotic management. Fever was positively associated with appropriate management (crude odds ratio [cOR], 4.33; 95% CI, 1.01–18.6; P=0.048). In contrast, abdominal pain or tenderness and signs of dehydration were negatively associated with appropriate management, with cORs of 0.43 (95% CI, 0.25–0.75; P=0.003) and 0.43 (95% CI, 0.21–0.87; P=0.019), respectively. Patient age ≥40 years, underlying medical conditions, and the presence of nausea or vomiting were not significantly associated with appropriate antibiotic management, with cORs of 0.83 (P=0.443), 1.15 (P=0.556), and 0.73 (P=0.298), respectively. Regarding physician-related factors, physician age was not significantly associated with appropriate management. However, physician specialty was significantly associated with antibiotic management, as residents were less likely to prescribe antibiotics appropriately than GPs (cOR, 0.44; P<0.001). In addition, antibiotic prescribing in the ED was more likely to be inappropriate than in the student PCU (cOR, 0.32; P=0.006). Pop-up insertion was not associated with improved appropriate antibiotic management (cOR, 0.95; P=0.827) (Table 3).
Table 3. Factors associated with appropriate antibiotic management.
| Factors | Crude odds ratio (95% CI) | P-value | |
|---|---|---|---|
| Sex, male | 1.03 (0.63–1.67) | 0.919 | |
| Medical conditions | 1.15 (0.72–1.85) | 0.556 | |
| Fevera | 4.33 (1.01–18.6) | 0.048 | |
| Nausea or vomiting | 0.73 (0.41–1.31) | 0.298 | |
| Abdominal pain/tendernessa | 0.43 (0.25–0.75) | 0.003 | |
| Signs of dehydrationa | 0.43 (0.21–0.87) | 0.019 | |
| Age ≥40 years | 0.83 (0.52–1.33) | 0.443 | |
| Pop-up insertion | 0.95 (0.62–1.46) | 0.827 | |
| Physician age ≥30 years | 0.81 (0.42–1.56) | 0.53 | |
| Site of visit (Student PCU as reference) | |||
| Emergency departmenta | 0.32 (0.14–0.72) | 0.006 | |
| Type of physician (GP as reference) | |||
| Residenta | 0.44 (0.27–0.71) | <0.001 | |
| Other specialists | 0.63 (0.19–2.05) | 0.439 | |
aStatistically significant factor.
CI, confidence interval; PCU, primary care unit; GP, general practitioner.
After adjustment for fever, site of visit, abdominal pain or tenderness, physician type, and signs of dehydration, only physician type remained independently associated with appropriate antibiotic management. Residents were less likely than GPs to prescribe antibiotics appropriately, with an adjusted odds ratio (aOR) of 0.38 (95% CI, 0.15–0.85; P=0.020) (Table 4).
Table 4. Factors independently associated with appropriate antibiotic management after adjustment for fever, site of visit, abdominal pain or tenderness, physician type, and signs of dehydration.
| Factors | Adjusted odds ratio (95% CI) | P-value | |
|---|---|---|---|
| Fever | 2.89 (0.57–14.8) | 0.202 | |
| Abdominal pain/tenderness | 0.61 (0.27–1.43) | 0.257 | |
| Signs of dehydration | 0.52 (0.25–1.09) | 0.083 | |
| Site of visit (Student PCU as reference) | 0.416 | ||
| Emergency department | 0.64 (0.22–1.87) | ||
| Type of physician (GP as reference) | 0.020 | ||
| Residenta | 0.38 (0.15–0.85) | ||
aStatistically significant factor.
CI, confidence interval; PCU, primary care unit; GP, general practitioner.
3. Factors associated with antibiotic prescription
Univariate analysis demonstrated that patients presenting with fever, mucous or bloody diarrhea, abdominal pain or tenderness, and signs of dehydration were more likely to receive antibiotic treatment, with cORs of 12.1 (95% CI, 4.50–32.6; P<0.001), 8.14 (95% CI, 2.19–30.2; P=0.002), 1.58 (95% CI, 1.00–2.50; P=0.052), and 2.30 (95% CI, 1.21–4.37; P=0.011), respectively. In contrast, no significant relationship was observed for age ≥40 years, underlying medical conditions, or nausea/vomiting. Among provider-related characteristics, physician age did not influence prescribing behavior. However, compared with GPs, antibiotic use was more frequent among residents and non-internist specialists, with cORs of 2.19 (P<0.001) and 3.84 (P=0.015), respectively. Encounters in the ED were also associated with higher rates of antibiotic prescription than those in the student PCU (cOR, 3.39; P=0.001) (Supplementary Table 1).
Multivariable analysis adjusting for relevant clinical and physician-related factors identified fever as an independent predictor of antibiotic prescription (aOR, 14.6; 95% CI, 2.54–83.8; P=0.003). In addition, resident physicians remained significantly more likely than GPs to prescribe antibiotics (aOR, 3.08; 95% CI, 1.04–9.09; P=0.042) (Supplementary Table 2).
Discussion
This study did not demonstrate the effectiveness of implementing a clinical pathway pop-up window within a computer-based prescribing system as an antibiotic stewardship intervention for acute diarrhea. The overall rates of antibiotic prescribing and appropriate antibiotic management were similar between the pre-intervention and post-intervention groups. Furthermore, segmented regression analysis showed no significant changes in the trends of antibiotic prescribing or appropriate antibiotic management following implementation of the intervention. In multivariable analysis, fever and management by resident physicians were significantly associated with increased odds of antibiotic prescription. In addition, management by resident physicians was associated with inappropriate antibiotic management.
Amid the growing threat of antimicrobial resistance, antibiotic stewardship has become an essential strategy to reduce inappropriate antibiotic use. Although most stewardship efforts have traditionally focused on inpatient settings, the majority of antibiotic prescriptions are issued in outpatient care [18]. Previous studies have demonstrated that multifaceted outpatient stewardship interventions can effectively reduce unnecessary antibiotic prescribing for acute diarrhea [11,19]. In contrast, a recent study from Bangladesh evaluating electronic clinical decision support for acute diarrhea did not demonstrate a significant reduction in antibiotic prescribing [20]. Similarly, our study found that implementing a clinical pathway pop-up within a computerized prescribing system did not significantly affect antibiotic prescribing behavior. One possible explanation for the lack of effectiveness is low physician adherence to the pop-up system. Many prescriptions either bypassed the pop-up without a documented indication or included indications that were inconsistent with the corresponding medical records, suggesting limited engagement with the intervention. In addition, physicians could proceed directly to the prescription page after selecting predefined indications, which may have further reduced the intervention’s impact. Although this intervention previously reduced antibiotic prescribing for URI in our setting [14], the lack of effect observed in the present study may reflect adaptation to or desensitization toward the pop-up system.
Antibiotic stewardship in acute diarrhea in Thailand may be more challenging than in URI. A substantial proportion of acute diarrhea cases in Thailand are attributable to bacterial pathogens, including Salmonella, Escherichia coli, Plesiomonas shigelloides, Vibrio species, and Campylobacter species [21,22]. This differs from URI, where clinicians can often emphasize that the illness is viral and that antibiotics are therefore unnecessary. In contrast, the clinical message in acute diarrhea is more nuanced: although the infection may be bacterial, most cases in immunocompetent patients with non-severe disease are self-limited and can resolve without antibiotic therapy. This distinction may create uncertainty for both physicians and patients, potentially widening the gap between guideline recommendations and prescribing behavior, and increasing the likelihood that antibiotics are requested or prescribed despite limited clinical benefit. Achieving an appropriate balance between restrictive stewardship strategies and physician autonomy is an important factor for long-term success and sustained cooperation [20]. Therefore, effective interventions likely require a multifaceted approach, including patient and prescriber education, development of locally adapted guidelines, monitoring of prescribing metrics, and provision of direct educational and constructive feedback to prescribers. Integration of these strategies into hospital policy and ED practice may help reduce unnecessary antibiotic prescribing [11].
Factors associated with antibiotic prescribing in this study can be broadly categorized into patient-related factors and healthcare system factors. Among patient-related factors, fever remained independently associated with antibiotic prescription in multivariable analysis, consistent with findings from a prior study conducted in the ED of a university hospital in Lebanon [23]. In addition, markers of disease severity—including signs of dehydration, tenesmus, and the need for hospital admission—were also significantly associated with antibiotic use [8,23,24]. These findings are in alignment with international guideline recommendations, which support antibiotic therapy in patients with more severe clinical presentations [17]. Regarding healthcare system factors, encounters with residents or non-internist specialists was significantly associated with higher odds of antibiotic prescribing. In addition, resident physicians were significantly associated with higher rates of inappropriate antibiotic management in the multivariable analysis, consistent with findings from a previous study conducted at King Chulalongkorn Memorial Hospital, Thailand [8]. The higher prescribing rates observed in these groups may reflect the need for rapid symptom control in acute care settings and time constraints associated with high patient volumes, which may limit adherence to guideline-based management. Other contributing factors may include limited motivation for education, perceived pressure to satisfy patient expectations, and insufficient experience in the management of acute diarrhea [8,14].
Our study has several limitations. First, given the quasi-experimental design, the quality and completeness of medical records could not be controlled. When physicians did not document specific clinical features, these variables were treated as missing data, which may have attenuated associations between certain factors and antibiotic prescribing. Second, the pop-up intervention was limited to the most commonly used antibiotics in our setting (norfloxacin and ciprofloxacin). Prescriptions for alternative regimens—such as cefixime, azithromycin, or co-trimoxazole—did not trigger the pop-up, potentially reducing the overall effect of the intervention. However, the frequency of these alternative prescriptions was low and is unlikely to have significantly influenced the primary outcome. Moreover, a subgroup analysis including only patients who received norfloxacin or ciprofloxacin showed results consistent with the main analysis. Third, the majority of patients were managed in the ED, reflecting local healthcare utilization patterns where access to emergency services is relatively convenient, even for non-urgent conditions. This may limit the generalizability of the findings to other settings. Fourth, as this study was conducted in an outpatient setting, microbiological data were not routinely available because of practical constraints. Finally, we did not implement additional interventions—such as physician and patient education, audit and feedback, or workflow modifications—which may enhance the effectiveness of clinical decision support tools. Moreover, Longer-term follow-up is necessary to evaluate the sustained effectiveness of the clinical pathway pop-up and its impact on prescribing behavior. Despite the fact that this study had negative outcomes but this study would provide information for establishing further modality to reduce antibiotic prescription rate and increase rational antibiotic drug prescription rate leading to hospital’s policy in antibiotic stewardship in acute diarrhea.
Insertion of a clinical pathway pop-up within a computerized prescribing system did not reduce overall antibiotic prescribing or improve the appropriateness of antibiotic use. After adjustment, fever and care by residents remained independently associated with antibiotic prescribing. These findings suggest that multimodal interventions beyond clinical decision support should be implemented in combination to effectively reduce inappropriate antibiotic use and mitigate the global threat of antimicrobial resistance.
Footnotes
Funding: This study was funded by Khon Kaen University (Grant No. IN63332).
Conflict of Interest: No conflict of interest.
Data sharing statement: The data that support the findings of this study are available from the corresponding author, upon reasonable request.
Disclosure of Artificial Intelligence (AI)-Assisted Technology: During the preparation of this work, the authors used ChatGPT (version 5.4) for writing assistance and grammatical editing. The authors subsequently reviewed and revised the content as necessary and take full responsibility for the final manuscript.
- Conceptualization: WS.
- Data curation: KK, NL.
- Formal analysis: WS.
- Investigation: NL, WS.
- Methodology: WS.
- Resources: NL, WS.
- Project administration: WS.
- Software: WS.
- Supervision: WS.
- Validation: WS.
- Visualization: WS.
- Writing - original draft: KK, WS.
- Writing - review & editing: WS.
SUPPLEMENTARY MATERIALS
Factors associated with antibiotics prescription
Factors independently associated with antibiotic prescription after adjustment for fever, mucous or bloody diarrhea, site of visit, abdominal pain or tenderness, physician type, and signs of dehydration
References
- 1.Aslam B, Wang W, Arshad MI, Khurshid M, Muzammil S, Rasool MH, Nisar MA, Alvi RF, Aslam MA, Qamar MU, Salamat MKF, Baloch Z. Antibiotic resistance: a rundown of a global crisis. Infect Drug Resist. 2018;11:1645–1658. doi: 10.2147/IDR.S173867. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Pumart P, Phodha T, Thamlikitkul V, Riewpaiboon A, Prakongsai P, Limwattananon S. Health and economic impacts of antimicrobial resistance in Thailand. J Health Serv Res Policy. 2012;6:352–360. [Google Scholar]
- 3.Kongnakorn T, Tichy E, Kengkla K, Kanokwanvimol N, Suthipinijtham P, Phuripakathorn C, Al Taie A. Economic burden of antimicrobial resistance and inappropriate empiric treatment in Thailand. Antimicrob Steward Healthc Epidemiol. 2023;3:e109. doi: 10.1017/ash.2023.169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Roca I, Akova M, Baquero F, Carlet J, Cavaleri M, Coenen S, Cohen J, Findlay D, Gyssens I, Heuer OE, Kahlmeter G, Kruse H, Laxminarayan R, Liébana E, López-Cerero L, MacGowan A, Martins M, Rodríguez-Baño J, Rolain JM, Segovia C, Sigauque B, Tacconelli E, Wellington E, Vila J. The global threat of antimicrobial resistance: science for intervention. New Microbes New Infect. 2015;6:22–29. doi: 10.1016/j.nmni.2015.02.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Ministry of Public Health. Bureau of Epidemiology Department of Disease Control. 2566 annual report. Epidemiological surveillance report 2023. [Accessed February 27 2026]. Available at: https://apps-doe.moph.go.th/boeeng/annual/Annual/Annual%20Report%202023.pdf.
- 6.Tulu S, Tadesse T, Alemayehu Gube A. Assessment of antibiotic utilization pattern in treatment of acute diarrhoea diseases in Bishoftu general hospital, Oromia Ethiopia. Adv Med. 2018;2018:2376825. doi: 10.1155/2018/2376825. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kotwani A, Chaudhury RR, Holloway K. Antibiotic-prescribing practices of primary care prescribers for acute diarrhea in New Delhi, India. Value Health. 2012;15(1 Suppl):S116–S119. doi: 10.1016/j.jval.2011.11.008. [DOI] [PubMed] [Google Scholar]
- 8.Supcharassaeng S, Suankratay C. Antibiotic prescription for adults with acute diarrhea at King Chulalongkorn Memorial Hospital, Thailand. J Med Assoc Thai. 2011;94:545–550. [PubMed] [Google Scholar]
- 9.AMR Thailand. Thailand’s national strategic plan on antimicrobial resistance 2017-2021. [Accessed February 27 2026]. Available at: https://amrthailand.net.
- 10.Gutiérrez G, Guiscafré H, Bronfman M, Walsh J, Martínez H, Muñoz O. Changing physician prescribing patterns: evaluation of an educational strategy for acute diarrhea in Mexico City. Med Care. 1994;32:436–446. [PubMed] [Google Scholar]
- 11.Boonyasiri A, Thamlikitkul V. Effectiveness of multifaceted interventions on rational use of antibiotics for patients with upper respiratory tract infections and acute diarrhea. J Med Assoc Thai. 2014;97(Suppl 3):S13–S19. [PubMed] [Google Scholar]
- 12.Madran B, Keske Ş, Uzun S, Taymaz T, Bakır E, Bozkurt İ, Ergönül Ö. Effectiveness of clinical pathway for upper respiratory tract infections in emergency department. Int J Infect Dis. 2019;83:154–159. doi: 10.1016/j.ijid.2019.04.022. [DOI] [PubMed] [Google Scholar]
- 13.Jenkins TC, Irwin A, Coombs L, Dealleaume L, Ross SE, Rozwadowski J, Webster B, Dickinson LM, Sabel AL, Mackenzie TD, West DR, Price CS. Effects of clinical pathways for common outpatient infections on antibiotic prescribing. Am J Med. 2013;126:327–35.e12. doi: 10.1016/j.amjmed.2012.10.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Sribenjalux W, Larbsida N, Khamsai S, Panyapornsakul B, Deawtrakulchai P, Meesing A. Insertion of a clinical pathway pop-up window into a computer-based prescription system: a method to promote antibiotic stewardship in upper respiratory tract infection. Antibiotics (Basel) 2021;10:1479. doi: 10.3390/antibiotics10121479. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Baur D, Gladstone BP, Burkert F, Carrara E, Foschi F, Döbele S, Tacconelli E. Effect of antibiotic stewardship on the incidence of infection and colonisation with antibiotic-resistant bacteria and Clostridium difficile infection: a systematic review and meta-analysis. Lancet Infect Dis. 2017;17:990–1001. doi: 10.1016/S1473-3099(17)30325-0. [DOI] [PubMed] [Google Scholar]
- 16.Department of Disease Control, Ministry of Public Health of Thailand and Gastroenterological Association of Thailand. Clinical practice guideline for acute diarrhea in adult. [Accessed 4 March 2026]. Available at: https://www.gastrothai.net/th/guideline-detail.php?content_id=33.
- 17.Shane AL, Mody RK, Crump JA, Tarr PI, Steiner TS, Kotloff K, Langley JM, Wanke C, Warren CA, Cheng AC, Cantey J, Pickering LK. 2017 Infectious Diseases Society of America clinical practice guidelines for the diagnosis and management of infectious diarrhea. Clin Infect Dis. 2017;65:e45–e80. doi: 10.1093/cid/cix669. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Sayood S, Durkin MJ. The challenge of outpatient antibiotic stewardship. JAMA Netw Open. 2023;6:e2312996. doi: 10.1001/jamanetworkopen.2023.12996. [DOI] [PubMed] [Google Scholar]
- 19.Gutiérrez G, Guiscafré H, Bronfman M, Walsh J, Martínez H, Muñoz O. Changing physician prescribing patterns: evaluation of an educational strategy for acute diarrhea in Mexico City. Med Care. 1994;32:436–446. [PubMed] [Google Scholar]
- 20.Nelson EJ, Khan AI, Keita AM, Brintz BJ, Keita Y, Sanogo D, Islam MT, Khan ZH, Rashid MM, Nasrin D, Watt MH, Ahmed SM, Haaland B, Pavia AT, Levine AC, Chao DL, Kotloff KL, Qadri F, Sow SO, Leung DT. Improving antibiotic stewardship for diarrheal disease with probability-based electronic clinical decision support: a randomized crossover trial. JAMA Pediatr. 2022;176:973–979. doi: 10.1001/jamapediatrics.2022.2535. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Bodhidatta L, Anuras S, Sornsakrin S, Suksawad U, Serichantalergs O, Srijan A, Sethabutr O, Mason CJ. Epidemiology and etiology of traveler’s diarrhea in Bangkok, Thailand, a case-control study. Trop Dis Travel Med Vaccines. 2019;5:9. doi: 10.1186/s40794-019-0085-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Treeprasertsuka S, Thepsuthammaratb K, Kitsahawongc B, Phaosawasdi K. Acute diarrhea, a significant burden to Thailand’s universal health care system: a nationwide database. Asian Biomed. 2016;10(Suppl 1):S23–S30. [Google Scholar]
- 23.Jabak SJ, Kawam L, El Mokahal A, Sharara AI. Management of acute diarrhea in the emergency department of a tertiary care university medical center. J Int Med Res. 2022;50:3000605221115385. doi: 10.1177/03000605221115385. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Lee HJ, Park KH, Park DA, Park J, Bang BW, Lee SS, Lee EJ, Kim YJ, Hong SK, Kim YR. Prescription of antibiotics for adults with acute infectious diarrhea in Korea: a population-based study. Infect Chemother. 2019;51:295–304. doi: 10.3947/ic.2019.51.3.295. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Factors associated with antibiotics prescription
Factors independently associated with antibiotic prescription after adjustment for fever, mucous or bloody diarrhea, site of visit, abdominal pain or tenderness, physician type, and signs of dehydration