The impact of the quality improvement program on Clostridioides difficile infection management: a quasi-experimental study
Article information
Abstract
Background/Aims
Clostridioides difficile infection (CDI) is a major cause of nosocomial diarrhea. This study aimed to implement a quality improvement program to expedite proper CDI treatment, including discontinuing laxatives and associated antibiotics.
Methods
Stool test results positive for CDI were automatically sent via text message to the quality improvement team, specialists in CDI management. The quality improvement team played an advisory role in this treatment. The outcome of this study was the competency of CDI treatment within 24 hours of stool test reporting. Competency was investigated using 3 different models: Model 1, initiation of CDI treatment within 24 hours of positive stool test report; Model 2, Model 1 criteria met with no concurrent laxative use; and Model 3, Model 2 criteria met with no concurrent associated antibiotics. Competency rates were compared between pre- and post-intervention periods (1 year each). Analyses were performed for inpatients with CDI.
Results
In total, 310 inpatients with CDI (129 pre-intervention, 181 post-intervention) were included in this study. The rates of competency for Model 1 (85.3% vs. 95.6%, P=0.006), Model 2 (81.4% vs. 92.3%, P=0.004), and Model 3 (35.7% vs. 56.4%, P<0.001) in the post-intervention group were higher to those in the pre-intervention group.
Conclusions
Quality improvement program enhanced the quality of CDI treatment in terms of prompt treatment and discontinuation of concomitant laxatives and associated antibiotics. (cris.nih.go.kr; KCT0005892)
INTRODUCTION
Clostridioides difficile infection (CDI) is a major cause of nosocomial diarrhea, leading to worsened underlying diseases, prolonged hospital stays, increased mortality, and increased medical costs [1]. Therefore, prompt diagnosis and appropriate treatment are essential for effective CDI management [2]. However, diagnosing CDI based solely on symptoms and laboratory tests can be challenging [3,4]. No gold standard diagnostic tool for CDI exists. Moreover, several confounding factors are present for CDI, such as medications or patient condition, which can cause diarrhea [5,6].
While gastroenterologists and infectious disease physicians are considered CDI management specialists, the infection can affect patients beyond these specialties [7,8]. Risk factors for CDI include antibiotic use, inpatient status (hospitalization), older age, compromised immune systems, and dysbiosis [8-10]. Although no prior studies have directly investigated the quality of CDI management based on physicians’ specialty, one study showed lower awareness of CDI among non-internal medicine doctors compared with their internal medicine colleagues [11].
Various quality improvement programs (QIP) have been developed to enhance CDI care [12-15], with most previous studies focusing on reducing unnecessary CDI testing. This study aimed to evaluate the feasibility of using a comprehensive QIP for CDI that addresses testing and improves the overall standard of CDI diagnosis and treatment.
METHODS
1. Study Setting
This single-center, quasi-experimental study was conducted at Bucheon St. Mary’s Hospital, a 700-bed tertiary hospital in Korea. The hospital established a comprehensive QIP for inpatients with CDI in March 2021. A preparatory period (January–February 2021) involved recruiting healthcare professionals into the quality improvement team (QIT) and developing program details. Data for the QIP pre-implementation (pre-QIP) period were collected retrospectively over the 12 calendar months before QIP implementation (March 2020 to February 2021). Data for the QIP post-implementation (post-QIP) period were prospectively collected over 12 calendar months after QIP implementation (March 2021 to February 2022). Informed consent was obtained from participants for the post-QIP period. We excluded the following: (1) outpatients, (2) hospice care patients, (3) patients with inflammatory bowel disease, and (4) those who refused participation in the post-QIP group.
2. QIP and CVR
The QIT comprised attending physicians from gastroenterology and infectious diseases, with a registered nurse as the study coordinator. They provided comprehensive guidance on both CDI diagnosis and management. Initially, a critical value report (CVR) system was adopted for the QIP. Positive CDI stool tests were text-messaged to the QIT and displayed in electronic medical records. To avoid alert fatigue, prescribers who ordered the stool test were not messaged. Stool tests were performed at the discretion of the prescriber. They included enzyme immunoassay (EIA) (Tox A/B Quik Chek; TECHLAB Inc., Blacksburg, VA, USA), simultaneous EIA and glutamate dehydrogenase (GDH) (C. Diff Quik Complete; TECHLAB Inc.), and polymerase chain reaction (PCR) (Xpert C. difficile; Cepheid, Sunnyvale, CA, USA).
During both the pre-QIP and post-QIP period, a positive EIA toxin test or PCR met the criteria for CDI diagnosis. The combination of GDH and toxin test was introduced during the post-QIP period in July 2021. Following this change, a diagnosis of CDI for cases that were toxin-negative but GDH-positive required a positive PCR test result [3]. In all instances where GDH yielded a positive result but the toxin test was negative, the QIT recommended that the doctor order a PCR test. CDI was defined as having 3 or more loose stools per 24 hours, along with a positive result from the stool study (Fig. 1) [3].
Definition of Clostridioides difficile infection. aStool study criteria were considered met when 1 of the 3 conditions was satisfied; bCriterion 3 was added after July 2021, at which point the Toxin test was replaced. by a combination Toxin/GDH test. PCR, polymerase chain reaction; GDH, glutamate dehydrogenase.
Initially, the physician who ordered the stool test (resident or attending) determined the treatment plan. The QIT evaluated the patient’s condition within 24 hours of CVR dispatch. When the treatment regimen was not prescribed, the QIT advised the patient’s doctor about the start of the treatment and the regimen. Further, when the treatment regimen was not properly adopted, the QIT advised correction of the treatment regimen. The QIT always advised the prescribers on the general management of CDI, such as discontinuing causative antibiotics and laxatives. When a treatment regimen was initiated for false positive results or colonization, it was advised to be stopped. A false positive was defined as either (1) GDH (+)/toxin (–)/PCR (–) results or (2) a number of diarrheal stools less than 3 per day. In all instances, the QIT evaluated the causes of diarrhea. During treatment, QIT involved treatment escalation, including fecal microbiota transplantation, in case of worsening symptoms [16,17].
3. Outcome
The outcome of this study was the competency of CDI treatment within 24 hours of stool test reporting through CVR, which met the CDI diagnostic criteria. Complete competency for CDI treatment was defined when all the following were met: (1) proper CDI treatment, (2) no concomitant use of laxatives, and (3) no concomitant use of associated antibiotics. Appropriate treatments included oral metronidazole or vancomycin for mild-to-moderate CDIs, oral vancomycin for severe CDIs, and high-dose oral vancomycin and intravenous metronidazole for fulminant CDIs [4]. CDI severity has been presented in a study elsewhere [4,18,19]. Severe CDI was defined as white blood cell count ≥15,000/μL or serum creatinine ≥ 1.5 mg/dL [4]. Fidaxomicin was not considered for CDI management since it was not available in Korea. Inappropriate CDI treatment after 24 hours of CVR dispatch was defined as: (1) use of intravenous metronidazole rather than oral metronidazole for mild-to-moderate CDIs; (2) use of oral metronidazole for severe CDIs; and (3) absence of a treatment regimen. In Korea, oral metronidazole is recognized as a suitable treatment for mild-to-moderate CDI because of its low prevalence in hypervirulent C. difficile strains [20]. CDI recurrence was defined as the recurrence of diarrhea after resolution and a positive stool test result within 8 weeks of CDI’s 1st episode. This study was approved by the Institutional Review Board of Bucheon St. Mary’s Hospital (No. HC20OISI0075) and registered at cris.nih.go.kr (KCT0005892). This study has been conducted in accordance with the Declaration of Helsinki. Informed consent was obtained from participants.
4. Statistical Analysis
We analyzed data from patients with at least 1 positive stool test for toxins, PCR, or GDH. Data for continuous and dichotomous variables were presented as mean±standard deviation and percentages, respectively, where appropriate. Demographic characteristics and outcomes of the quality measures were compared between the 2 groups (pre-QIP and post-QIP) using Student t test, chi-square test, or Fisher exact test where appropriate. Competency was investigated using 3 different models: Model 1, initiation of proper CDI treatment within 24 hours of stool test report (CVR dispatch); Model 2, fulfilling criteria for Model 1 and no concomitant use of laxatives; and Model 3, fulfilling criteria for Model 2 and no concomitant use of causative antibiotics. A two-sided P-value of less than 0.05 was considered statistically significant. Statistical analyses were performed using the R software (version 4.2.3; R Foundation for Statistical Computing, Vienna, Austria).
RESULTS
A total of 455 stool tests for CDI were analyzed during the 24-month study period. Of these, 103 patients were diagnosed with false-positive CDI results. The number of false positives or colonization was 39 in the pre-QIP and 64 in the post-QIP groups, respectively. The reasons for the false positive results include the absence of diarrhea, self-limiting diarrhea, and stool test results, which were insufficient for CDI diagnosis. The treatment regimen for CDI, which was considered overtreatment, was prescribed to 29 patients in the pre-QIP group and 10 in the post-QIP group (74.4% vs. 15.6%; P<0.001). In total, 352 patients were diagnosed with CDI. After excluding 42 patients, 310 patients were included in the analysis (Fig. 2).
Study schema and enrollment of patients. CDI, Clostridioides difficile infection.
1. Demographics and Baseline Characteristics
The demographic characteristics are presented in Table 1. Sex was comparable between the 2 groups. Age (pre-QIP 75.2 vs. post-QIP 71.4; P=0.015) and the Charlson Comorbidity Index score (pre-QIP 5.5 vs. post-QIP 4.7; P=0.001) were higher in the pre-QIP group than in the post-QIP group. Antibiotics were used before CDI diagnosis in 93.1% and 91.5% of patients in the pre- and post-QIP periods, respectively, within 2 months of CDI diagnosis (P=0.620). The most common reason for antibiotic use in both groups was pneumonia. Internal medicine departments ordered most stool tests (223/310, 71.9%). Death rates within 2 months of CDI diagnosis were 15.5% and 6.1% in the pre- and post-QIP groups, respectively (P=0.006).
Demographic Characteristics and Risk Factors for CDI of the Study Subjects
2. Variables Related to CDI
The variables related to CDI are presented in Table 2. The severity of CDI was comparable between the 2 groups. However, diagnostic methods differed between the 2 groups. Additionally, the number of CDI episodes and initial treatment regimens differed between the 2 groups. In the pre-QIP group, the treatment regimen was not prescribed to 7.0% of the patients. Notably, no patient in the post-QIP group was left without a treatment regimen.
Variables Related to CDI
3. Quality Improvement
Table 3 compares the competency of CDI treatment assessed 24 hours after the stool test report. The rate of CDI treatment competency was higher in the post-QIP group than in the pre-QIP group in all 3 models (Model 1: pre-QIP 84.5% vs. post-QIP 93.9%, P =0.006; Model 2: pre-QIP 81.4% vs. post-QIP 92.3%, P=0.004; Model 3: pre-QIP 35.7% vs. post-QIP 56.4%, P<0.001). Concomitant laxative use was higher in the pre-QIP group than in the post-QIP group (pre-QIP 8.5% vs. post-QIP 1.7%; P=0.005). Associated antibiotic use was also lower in the post-QIP group than in the pre-QIP group (pre-QIP 56.6% vs. post-QIP 41.4%; P=0.008). At 24 hours after the stool test report, the treatment regimen was not prescribed in 4.4% of post-QIP cases, significantly lower than the pre-QIP rate of 10.9% (P=0.030). Details of the QIP recommendations are as follows: improper regimen changes (n=11); initiation of a treatment regimen for CDI (n=37); cessation for associated antibiotics for CDI or laxative (n=22). Notably, prescribers adopted 70 out of the 80 recommendations made by the QIT. Half of these recommendations involved initiating CDI treatment in patients who had not yet received it.
Competency of CDI Treatment
DISCUSSION
In this study, we investigated the effectiveness of a QIP in improving CDI management. We found that the post-QIP group demonstrated significantly higher competency in CDI treatment than the pre-QIP group across all 3 competency models.
To our knowledge, the impact of QIP on CDI treatment has not been investigated in previous studies. During the post-QIP period, stool test results were collected for a year using a CVR system, typically used for critical findings, such as cancer diagnosis and life-threatening laboratory findings [21,22]. Notably, this was the first reported use of a CVR system for CDI treatment.
Effective CDI treatment includes accurate diagnosis, appropriate antibiotic regimens, and general management strategies, such as discontinuing laxatives and associated antibiotics. Physician knowledge of CDI depends on their specialty [11]. Our QIP aimed to address this by guiding various prescribers (often non-specialists in CDI) who ordered stool tests, ultimately improving treatment competency. We evaluated treatment competency 24 hours after the stool test report. In this study, various competency models have been adopted with increasing levels of stringency.
Previous studies reported a considerable proportion of patients receiving laxatives also undergoing C. difficile testing [11,23]. In our analysis, competency rates for Models 1 and 2 exceeded 90% in the post-QIP group. However, the competency rate in Model 3, which considers concomitant antibiotic use, was only 56.4% post-QIP because of the high rate of continued antibiotic use. Although the concomitant use of associated antibiotics was lower following the QIP, the rate 24 hours after the stool test report was still as high as 43.9%. This persistence may be due to the need for antibiotics to treat the original infection even when CDI is diagnosed.
During the post-QIP period, approximately three-fourths of our hospital’s CDI patients were treated by non-specialists(physicians other than gastroenterologists or infectious disease doctors). In the same period, prescribers adopted 70 of the 80 QIT recommendations for inpatient CDI treatment. This suggests that QIP effectively supported physicians who were not CDI specialists, thereby increasing the quality of CDI treatment.
Our QIP focused on achieving prompt treatment initiation with QIT support. However, the QIP, while not focused on treatment initiation, did reduce unnecessary treatment for colonization or false positives. Diagnosing CDI can be complex [3,4]. Overdiagnosis can lead to unnecessary antibiotic treatment, thereby increasing healthcare costs [24]. The rate of overtreatment, in which the treatment regimen was prescribed for false positives, was significantly higher in the pre-QIP group than in the post-QIP group (74.4% vs. 15.6%; P<0.001). In a previous study, a reduction in unnecessary antimicrobial administration was achieved by implementing antimicrobial stewardship programs [14].
In this study, metronidazole was used along with vancomycin for the initial treatment of CDI. When this study was conducted, Korea lacked national guidelines for CDI treatment, and oral metronidazole was a common choice [4,20]. The low prevalence of hypervirulent C. difficile strains in Korea (less than 5%) justifies this approach [25,26]. However, vancomycin use carries the risk of vancomycin-resistant Enterococcus infection [27].
This study had some limitations. First, the pre-QIP and post-QIP data were not collected identically. Pre-QIP data were retrospectively investigated, and some patients may not have been followed for treatment outcomes. During the pre-QIP period, 7% of patients did not receive CDI treatment, and their outcomes were not assessed. Second, the study was conducted at a single center, limiting the generalizability of the findings. A multicenter study would be necessary to confirm the effectiveness of the QIP in other settings.
In conclusion, the implementation of a comprehensive QIP for CDI management demonstrably enhanced the quality of CDI treatment by promoting prompt treatment initiation, reducing unnecessary laxative use, and encouraging discontinuation of causative antibiotics. Further research using multicenter designs is needed to confirm these findings and explore the broader applicability of the QIP.
Notes
Funding Source
This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (grant number: NRF-2021R1G1A1094049) and the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (grant number: HI19C0481, HC20C0099).
Conflict of Interest
No potential conflict of interest relevant to this article was reported.
Data Availability Statement
T he datasets in this study are not publicly available because of privacy policies. Please contact the corresponding author for data requests that are limited and reasonable.
Author Contributions
Conceptualization: Gweon TG, Kim CW, Cho YS. Data curation: Lee JY, Kim HS. Formal analysis: Lee JY, Kim HS. Funding acquisition: Gweon TG, Cho YS. Investigation: Choi JK, An YE. Methodology: Choi JK, Kim CW. Project administration: Kim TH. Resources: Kim HS, Cho YS. Software: Choi JK, Chang JH. Supervision: Gweon TG, Chang JH. Validation: Kim HS. Visualization: Chang JH, Kim TH. Writing - original draft: Lee JY, Choi JK. Writing - review & editing: Gweon TG, An YE, Kim HS, Chang JH, Kim TH, Kim CW, Cho YS. Approval of final manuscript: all authors.