An antimicrobial stewardship program (ASP) is considered a vital strategy to reduce antimicrobial resistance (AMR), which has rapidly spread worldwide.Reference Huttner, Harbarth and Carlet1 An ASP is a set of multidisciplinary activities designed to use antimicrobials in ways that ensure sustainable access to effective therapy for all who need it with the aim of obtaining optimal clinical outcomes, reducing potential adverse effects, and arresting the development of AMR.Reference Dyar, Huttner, Schouten and Pulcini2
According to a past survey, the ASP activity in most Korean hospitals is heavily dependent on the restrictive measures for designated antimicrobials, which is similar to a preauthorization-of-antibiotic use program in that prescription of certain antibiotics is restricted unless approval is granted.Reference Hwang and Kim3 Although its efficacy is well established, there are several drawbacks with this measure: only a small proportion of overall antibiotic use can be controlled,Reference Hwang and Kim3 and it may disturb and break down trust among physicians.Reference Barlam, Cosgrove and Abbo4,Reference Davey, Marwick and Scott5 Therefore, implementing another effective program, such as a prospective audit and feedback program, which is a cornerstone of ASP activity, should be considered.Reference Tamma, Avdic and Keenan6
To diversify and improve ASPs in Korean hospitals, secure and sufficient manpower is mandatory.Reference MacDougall and Polk7 Unfortunately, in most Korean hospitals, ASPs are carried out by only 1 or 2 infectious diseases specialists.Reference Kim, Lee and Moon8 According to the Infectious Disease Society of America’s guidelines, each hospital should have an antimicrobial stewardship team infectious diseases physician and clinical pharmacist with infectious diseases training, and both of these individuals should dedicate enough time for ASP activities.Reference Barlam, Cosgrove and Abbo4 However, an acceptable staffing standard for an ASP team is currently unclear.
To achieve structural and financial support from stakeholders, clarification about how much time is needed to perform essential ASP activities adequately, and how many full-time equivalents (FTEs) are needed, should be provided at the outset. Hence, we performed this study to estimate the human resources required for ASP activities in Korean hospitals.
Material and methods
Study design
We conducted a multicenter retrospective study in 8 hospitals in Korea. The participating hospitals had 295–1,337 beds and were located throughout the Korean peninsula (3 in Seoul, 2 in Busan, 1 in Daegu, 1 in Incheon, and 1 in Gyeonggi-do), and 7 of 8 were university-affiliated hospitals. The time required to perform ASP activities on all hospitalized patients under antibiotic therapy was estimated and converted into hours per week. The actual time spent on patient reviews of each ASP activity was measured with a small number of cases, then the total time was estimated by applying the determined times to a larger number of cases. The FTEs were measured according to labor laws in Korea (52 hours per week).
In this study, antibiotics were defined as medication with Anatomical Therapeutic Chemical class J01, which does not include antifungal agents or antituberculotic agents.
Definition of major ASP activities
Major ASP activities were suggested through the discussions among 4 investigators: S.Y. Park, H.-H. Chang, E.S. Kim, and B. Kim. The draft activities were reviewed and revised through discussions with 10 infectious diseases specialists in the study group. Finally, 10 major ASP activities were determined as follows:
-
(1) Appropriate antibiotics recommendations for patients with suspected infections without a proven site of infection or causative pathogens
-
(2) Appropriate antibiotics recommendations for patients with suspected infections without a proven site of infection but with causative pathogens
-
(3) Appropriate antibiotics recommendations for patients with suspected infections with a proven site of infection but without causative pathogens
-
(4) Appropriate antibiotics recommendations for patients with suspected infections with a proven site of infection and causative pathogens
-
(5) Review and approval of designated antibiotics under restrictive measures
-
(6) Avoidance of inappropriate combinations of antibiotic therapy
-
(7) Parenteral to oral conversion
-
(8) Review of therapy for appropriate duration, appropriate dosage, or drug–drug interaction
-
(9) Review of surgical prophylactic antibiotics
-
(10) Other.
The ‘appropriate antibiotic recommendation for patients with suspected infections without a proven site of infection or causative pathogens’ included empirical antibiotic use for sepsis and stopping antibiotics if there was no infectious disease that required antibiotic therapy.
Measuring time spent on each ASP activity
The time spent on patient review was measured in 4 of the participating hospitals. These hospitals had 734–1,337 beds, and all were university-affiliated hospitals. In total, 60 cases were selected randomly from patients with a request for formal infectious diseases consultation in each hospital from January 21, 2019, to January 26, 2019. Only 1 consultation per patient was included. One infectious diseases specialist in each hospital counted the number of ASP activities for each case (Supplementary Table 1 online). If 2 or more activities were observed in a single case, they were divided into ‘main’ and ‘additional’ activities according to the observer. Simultaneously, the time spent on patient review to perform the ‘main’ ASP activities was measured for each case and the median time required for each activity was calculated. The time spent on patient reviews included (1) review for medical records, (2) interview with patients, and (3) document antibiotic recommendations in medical records. If an ASP activity was observed in <5 cases as a ‘main’ activity, we allocated 10 minutes for it based on extrapolations from stewardship expert estimation. As for ‘additional’ activities, we allocated 2 minutes of stewardship expert estimation for each, regardless of the type of activity.Reference Echevarria, Groppi, Kelly, Morreale, Neuhauser and Roselle9
In addition, the following data were collected to assess the baseline characteristics of enrolled patients: age, sex, hospitalization duration, admission department, admission location, Charlson’s comorbidity index, positive culture results from sterile sites, and previous consultations with the infectious diseases department.
Measuring the number of each ASP activity for hospitalized patients per day
The number of ASP activities for hospitalized patients carried out per day was counted in all participating hospitals. Because a sample size of 1,574 produces a 2-sided 95% confidence interval with a width equal to 0.025 when the sample proportion is 0.500,Reference Newcombe10 we allocated 210 cases to 7 hospitals and 104 cases to 1 hospital with a smaller size than others. One infectious diseases specialist in each hospital randomly sampled the allocated number of cases from all hospitalized patients under systemic antibiotic therapy (per oral or parenteral routes) on January 23, 2019, using Excel 2016 (Microsoft, Redmond, WA) and counted the number of required ASP activities for each (Supplementary Table 1 online). Similar to the measurement time spent on patient reviews, if 2 or more activities were required for a single patient, they were divided into ‘main’ and ‘additional’ activities by the observer.
Estimation of FTE
The number of hospitalized patients and those undergoing antibiotic therapy were collected from each hospital for the estimation of FTE. Let M asp denote the median (interquartile range [IQR]) time spent on ‘main’ ASP activity (asp = 1 …. 10). The time spent on the review of a patient who underwent antibiotic therapy (T) was estimated as follows:
$${{\left( \eqlign{\eqalign{\mathop \sum \nolimits_{asp = 1}^{10} \left| {Masp*number\;of\;main\;ASP\;activity} \right| \cr+ \;2*number\;of\;additional\;ASP\;activity} }}\right)} \over {number\;of\;enrolled\;patients}}$$
Then, the unit was adjusted by 100 patients underwent antibiotic therapy (T100p) or 100 beds per day (T100b). T100b was calculated as follows:
$${{\hskip-4pt T\!*Number\,of\,hospitalized\,patients\,under\,antibiotic\,therapy} \over {Number\,of\,hospitalized\,patits\;}}*\!100$$
Finally, FTE was calculated based on the labor laws in Korea, which limit hours of work to a maximum of 52 hours per week. We defined a week as 5 working days. Times for vacations and sick leave were not considered in the calculation.
$${{T100p\;\left( {or\;T100b} \right)*5} \over {52*60}}$$
Statistical analysis
All statistical analyses were conducted using SPSS version 24.0 software for Windows (IBM, Armonk, NY). Continuous variables were analyzed using the Mann-Whitney U test. A 2-tailed P value of <.05 was considered statistically significant.
Ethics statements
The study protocol was approved by the Institutional Review Board of Hanyang University Seoul Hospital (IRB no. 2019-02-001) and each of the participating hospitals. The requirement for written informed consent from patients was waived due to the retrospective nature of the study.
Results
Measuring time spent on patient reviews
In total, 240 consultations were collected from 4 hospitals and 15 were excluded because they were not associated with ASP activities. Finally, 225 cases were analyzed and the baseline characteristics for patients are summarized in Table 1.
Table 1. Baseline Characteristics for Patients in the Study for Time Spent per Patient Review
Note. SD, standard deviation; IQR, interquartile range.
The mean patient age was 64.1, and 59.1% of the patients were male. The median hospitalization duration on point-prevalence survey days was 13 days (interquartile range [IQR], 6–28). Also, 44.0% of patients were admitted to medical departments and 81.3% were admitted to general wards. The mean Charlson’s comorbidity index score was 4.0. Among the 225 patients, 28.0% had positive culture results from sterile sites and 60.0% underwent consultation with the infectious diseases department.
The most frequently observed main ASP activity was ‘appropriate antibiotics recommendations for patients with suspected infections with a proven site of infection and causative pathogens,’ followed by ‘appropriate antibiotics recommendation for patients with suspected infections with a proven site of infection but without causative pathogens’ (Supplementary Table 2 online). Table 2 depicts the median time spent per patient review for each ASP activity. The median time spent ranged from 10 to 16 minutes, and there were no significant differences in time spent per patient review for each activity between patients in general wards and those in intensive care units. The most time-consuming activity was ‘parenteral to oral conversion’ (16 minutes), followed by ‘appropriate antibiotics recommendations for patients with suspected infections without a proven site of infection or causative pathogens’ (14 minutes), and ‘appropriate antibiotics recommendation for patients with suspected infections with a proven site of infection and causative pathogens’ (14 minutes). Fewer than 5 cases were observed for the following 4 activities, and 10 minutes were allocated for each: (1) avoidance of inappropriate combination antibiotic therapy, (2) review of therapy for appropriate duration, appropriate dosage, or drug–drug interaction, (3) review of surgical prophylactic antibiotics, and (4) other.
Table 2. Median Time Spent per Patient Review for Each Antimicrobial Stewardship Activity
Note. IQR, interquartile range; N/A, not available.
a The Mann-Whitney U test was performed for comparisons between patients in general wards and those in intensive care units.
b We allocated 10 minutes for the activities in which there were <5 cases.
Counting the ASP activities for hospitalized patients per day
In total, 1,534 patients who underwent antibiotic therapy were selected from 8 hospitals. In total, 64 hospitalized patients underwent antibiotic therapy at the smaller hospital on the surveillance day, and all cases were reviewed at the hospital.
Table 3 lists the observed counts for each ASP activity. The most commonly observed main activity was ‘review of surgical prophylactic antibiotics’ (32.7%), followed by ‘appropriate antibiotics recommendations for patients with suspected infections without a proven site of infection but without causative pathogens’ (28.6%) and ‘appropriate antibiotics recommendations for patients with suspected infections with a proven site of infection and causative pathogens’ (16.8%). A similar distribution was observed for patients in general wards. In comparison, ‘appropriate antibiotics recommendation for patients with suspected infections without a proven site of infection but without causative pathogens’ (37.4%) was most commonly observed in patients in intensive care units, followed by ‘appropriate antibiotics recommendation for patients with suspected infections with a proven site of infection and causative pathogens’ (23.7%), and ‘review of surgical prophylactic antibiotics’ (14.4%).
Table 3. The Number of Each Antimicrobial Stewardship Activity for Hospitalized Patients Observed Per Day
As for additional activity, a total of 1,518 activities were observed in 1,534 patients. ‘Review of therapy for appropriate duration, appropriate dosage, or drug–drug interaction’ (59.3%) were the most common additional activities, followed by ‘avoidance of inappropriate combinations of antibiotic therapy’ (24.2%), and ‘parenteral to oral conversion’ (12.2%). Similar distributions were observed for patients in general wards and intensive care units.
Estimation of FTE
The number of hospitalized patients in participating hospitals ranged from 580 to 1,289. Of these, 64–727 patients (32.5%–79.0%) underwent antibiotic therapy on the day of the point-prevalence survey (Table 4).
Table 4. The Number of Hospitalized Patients and Those Undergoing Antibiotic Therapy on the Day of Point Surveillance in Each Hospital
Note. ICU, intensive care unit.
Table 5 provides the estimated FTEs required for ASP activities for hospitalized patients calculated based on the measured and collected data. In our estimation, the personnel requirement was calculated as 1.04 FTEs (IQR, 0.85–1.22) per 100 beds for main activities and 1.20 FTEs (IQR, 1.02–1.38) per 100 beds for main and additional activities. When the unit was adjusted for 100 patients who underwent antibiotic therapy, the personnel requirement was calculated as 1.96 FTEs (IQR, 1.62–2.31) per 100 patients for main activities and 2.28 FTEs (IQR, 1.93–2.62) per 100 patients for main and additional activities.
Table 5. Estimated FTEs Required for Antimicrobial Stewardship Program Activities for Hospitalized Patients in Korean Hospitals
Note. FTE, full-time equivalent; IQR, interquartile range.
Discussion
To the best of our knowledge, this is the first study in Korea to estimate the human resource expenditure for ASP activities. Similar to our results, a study in the United States that measured actual time spent on ASP activities estimated that 1.0 FTE per 100 beds are needed to implement and manage a robust ASP.Reference Echevarria, Groppi, Kelly, Morreale, Neuhauser and Roselle9 The main result of the present study might be overstated because the study design focused on the ‘action’ of ASP, assuming the review of all hospitalized patients under antibiotic therapy. Although securing 1.20 FTEs (IQR, 1.02–1.38) per 100 beds for ASP activities seems difficult in the current situation, these findings are valuable for establishing an ASP policy in Korea, and the FTEs can be modified according to the available resources. For instance, the number of FTEs required for reviewing patients receiving broad-spectrum antibiotics (~50% of patients under antibiotic therapy) might be estimated at 6.02 FTEs (IQR, 5.10–6.92) per 1,000 beds; if the review were confined to antibiotics used for >1 week (~25% of patients under antibiotic therapy), and the personnel requirement might be estimated at 3.01 FTEs (IQR, 2.55–3.46) per 1,000 beds.Reference Kim, Hwang, Kim, Lee and Pai11–Reference Song Mi Moon13
The essential organizations that need to be established in hospitals to carry out the interventions suggested by the 7 Core Elements of the US Centers for Disease Control and Prevention may involve antimicrobial committees and ASP operational teams.Reference Barlam, Cosgrove and Abbo4 The role of an antimicrobial committee is to validate the strategy of ASP at an institutional level with regular meetings, whereas the role of an ASP operational team is to perform daily interventions in real clinical settings. The ASP team should include an infectious diseases physician, a clinical pharmacist (with infectious diseases training, if possible), a clinical microbiologist, an infection control specialist, and an information system specialist.14 In a typical ASP team, the key members are an infectious diseases physician who leads the team and is responsible for the implementation and evaluation of programs, and a clinical pharmacist who performs daily tasks related to ASP and supports the leader.Reference Apisarnthanarak, Kwa and Chiu15
Several studies pointed out that the lack of manpower is one of the main barriers to implementing an ASP.Reference Kim, Lee and Moon8,Reference Howard, Pulcini and Levy Hara16 Although human resources are an important factor for ASP operations, many hospitals in Asia do not have adequate personnel to make up a proper ASP team.Reference Levy Hara17 In Korea, the ASPs in hospitals are primarily carried out by 1 or 2 infectious diseases physicians, and the participation of other medical personnel such as clinical pharmacists and other specialists is still lacking.Reference Kim, Lee and Moon8 As a result, ASP activity in Korean hospitals has been highly dependent on the restrictive measures for designated antimicrobials, which covers only a small proportion of overall antibiotic use.Reference Kim, Lee and Moon8,Reference Kim, Kim, Kim and Pai18
An increase in FTEs of personnel participating in ASPs can predict a significant increase in the effectiveness of ASPs and the increased number of implemented ASPs in the hospital.Reference Maeda, Muraki and Kosaka19,Reference Doernberg, Abbo and Burdette20 A recent nationwide study about ASPs in Korean hospitals also demonstrated that the number of infectious diseases specialists and clinical pharmacists was associated with performing a higher number of antimicrobial intervention programs in a hospital.Reference Kim, Lee and Moon8 The recommended FTEs vary by country and range from 2 to 6 per 1,000 acute-care beds.Reference Pulcini, Morel and Tacconelli21 However, most recommendations are not evidence-based; rather, decisions are based on the healthcare system and the resources in each country.
To secure adequate human resources and improve the ASPs in Korean hospitals, government action with respect to investments for human resources and proper compensation for ASPs need to be supplemented. First, medical personnel with infectious diseases training should be fostered, and the most effective ASP team possible in the healthcare environment should be created. Although ASPs may be best led by infectious diseases physicians,Reference Barlam, Cosgrove and Abbo4,Reference Ostrowsky, Banerjee and Bonomo22 it is hard to depend on them due to the limited number of infectious diseases physicians in Korea: 274 infectious diseases specialists existed in 2019 and 15–20 new board-certified infectious diseases specialists are trained per year (data not shown). Therefore, inclusion of other staff, such as physicians with ASP training, clinical pharmacists, and nurse practitioners, could be considered as well. Second, an appropriate reward structure for ASP activities should be implemented in the current healthcare system in Korea. Because of the lack of compensation, Korean hospitals are reluctant to invest in implementing ASPs. One of the most desirable measures to promote hospital attention to the implementation of ASPs, might be the establishment of a healthcare fee for ASPs.Reference Kim, Lee and Moon8 Another measure might be institutionalizing the performing ASPs as a mandatory regulation for each hospital. In the United States, the Joint Commission has set up an ASP as a component of accreditation, and the Centers for Medicare and Medicaid Services has required hospitals to use ASPs.Reference Echevarria, Groppi, Kelly, Morreale, Neuhauser and Roselle9
This study has several potential limitations. First, the characteristics of patients selected for measuring the time spent on each ASP activity might be different from those who underwent antibiotic therapy in the hospital. Because the ASPs in most Korean hospitals are carried out by 1 or 2 infectious diseases physicians without a full-time staff member for the ASP, it is difficult to clearly divide the boundaries of ASP activities and infectious diseases consultation.Reference Kim, Lee and Moon8 In fact, the substantial proportion of infectious diseases consultations in Korean hospitals are associated with ASP activities (eg, the recommendation of proper antibiotic choice, route, dose, and duration), and the ‘restrictive measures for designated antimicrobials’ program is linked to infectious diseases consultation.Reference Hwang and Kim3,Reference Kim23 Nevertheless, the antibiotic choice for patients with antibiotic consultation requests might be more complex than that of other patients. Therefore, time spent on the review of a patient who underwent antibiotic therapy might be overestimated. Second, direct patient interview was included in the calculation of time spent on patient reviews. However, it might not be a standard practice of ASP activities in some hospital settings. Therefore, the applicability of our estimation to sites where practice ASP differently might be limited as well. Third, interobserver differences might exist. Although the training system of infectious disease specialists is similar across Korean hospitals, and the quality is controlled by the Korean Association of Internal Medicine, the classification of ASP activities and the practices for patient review might differ between investigators. Fourth, the study was conducted mainly in large hospitals and patients enrolled in the study might have more underlying comorbidities or be in more serious conditions compared to patients in smaller hospitals. Therefore, the results cannot be generalized to all hospitals. Finally, other core elements for ASP, such as education, tracking, or reporting, were not considered in this study.
In conclusion, the estimated human resources required for performing extensive ASP activities on all hospitalized patients undergoing antibiotic therapy in Korean hospitals were ~1.20 FTEs (IQR, 1.02–1.38) per 100 beds. This study only focused on the actions of ASPs, and further studies considering other core elements of ASPs are necessary to estimate more meticulous FTEs.
Acknowledgments
We acknowledge Da Young Kim, Jun-Won Seo, Eunjeong Heo, Hyung-Sook Kim, Hyuck Lee, Hojin Lee for their assistance with data collection.
Financial support
This study was supported by the research fund of The Korean Society of Infectious Diseases (2018); a grant from the Bio & Medical Technology Development Program of the National Research Foundation (NRF) and funded by the Korean government (MSIT grant no. 2019M3E5D1A01066063). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Conflicts of interest
All authors report no conflict of interest relevant to this study.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/ice.2020.1234
