Keywords:Anesthesiology; Feasibility Studies; Intensive Care Units; Workforce; Burnout

Abbreviations:ICU: Intensive Care Unit; APP: Advanced Practice Provider; NP: Nurse Practitioners; PA: Physician Assistant; CCM: Critical Care Medicine; FTE: Full Time Equivalent; SCCM: Society of Critical Care Medicine

Introduction

There is considerable variation amongst intensive care unit (ICU) staffing models, and a paucity of knowledge specifically on anesthesia critical care staffing [1]. When developing an ICU staffing model, one must consider multiple factors, such as the patient population, acuity, census, availability of non-physician advanced practice providers (APP) (Nurse Practitioners (NP) or Physician Assistants (PA)), the number of residents or fellow physicians, and the number of critical care medicine (CCM) physicians [2]. In addition, the work shifts including time off for the CCM physicians must be determined [3]. Further, compensation, either monetary or days off, may be given for the extra hours worked in the previous week. Due to these variables, it is not surprising that there is tremendous heterogeneity in staffing models and CCM physician work schedules. This heterogeneity makes surveying, categorizing, and comparing work shifts challenging.

Unlike operating room staffing, CCM physician shifts often involve evenings, nights, or weekends for several consecutive days. Increased workload coupled with patient acuity, bed shortages and staffing challenges, increase the days off needed for recovery and prevention of burnout though comes at a cost as “time is money” [4]. This “time” is a cost opportunity that not only allows for recovery from a work shift, but also time to develop professionally through academic pursuits, complete administrative responsibilities for an institution, or seek personal bliss. As a result of limited national surveys on staffing, CCM academic divisions and private groups struggle with defining a Full Time Equivalent (FTE) for a CCM physician and justifying their staffing model and days off when negotiating budgets with hospital or medical school administrators.

This feasibility study focused on determining if our novel data collection form would result in meaningful data collection to allow for staffing model comparisons and assessment of days off after a CCM physician work shift. As a result, leadership will be able to staff critical care services with sufficient resources that promote both patient care and physician wellbeing.

Methods

After meeting criteria for exemption from review by the University of Texas Medical Branch’s Institutional Review Board on September 14, 2022, for “Developing survey tool for benchmarking days off after working in critical care units” (IRB #22 0225), survey invitations were sent by email to professional colleagues in United States academic anesthesiology departments identified by investigators for this feasibility study without regard to geographic location. Procedures followed were in accordance with the ethical standards of the responsible committee and with the Helskinki Declaration of 1975, though no human or animal subjects were involved with this work. Final survey questions were developed through feedback and data from a pilot of six ICUs involving three states from the sites of the co-authors. Thirtythree contacts in academic anesthesiology departments were identified. Surveys were distributed on September 15, 2022, with one reminder before closing the survey on November 18, 2022.

A data collection form was completed for each ICU covered by the department. This two-page survey included two parts and an illustrative example to aid in understanding how to complete this novel data collection form. The first part included instructions and three “demographic” questions about the ICU including the name of the ICU, the number of total beds available, and if residents, fellows, and/or APPs were present, or if the CCM physician worked alone (Figure 1a). In addition, the time to complete the survey for each ICU was collected. On the second part, respondents completed the data collection form based on a typical week of ICU coverage (Figure 1b). For each CCM physician, an hourly table for weekdays and weekends was completed and the number of days off the week after ICU coverage was documented. Respondents were asked if additional monetary compensation above their base pay was provided for working the shift (yes/no response).

Given the survey population of anesthesiologist CCM physicians, another way to think of this question was to compare if they instead worked in the operating room for those shifts would their pay decrease (YES) or remain the same (NO). Respondents indicated if the CCM physician worked specific hours, if the physician could leave the facility (i.e., home call), and if the physician worked alone for the hours (physician only). The form did not ask respondents to categorize the work shifts. Since the data collection form was structured on an hourly basis, and ascertained if it was in house or on home call, the data allowed for identification of work shifts. Examples of possible shifts included weekday in house without weekday night or weekend duties, weekday in house without weekday night but with weekend home duties both day and night, or only working in house weekday nights. Final shifts were determined by responses. In reporting data by all work shifts and then specific work shifts, the survey also allows subcategories based on number of beds and type of other clinicians working with the CCM physician. Days off after a CCM work shift are reported as average, median, standard deviation, and 25th and 75th percentiles for shifts with at least five responses.

Results

Thirty-three academic anesthesiology departments received a data collection form. Nineteen of 33 programs completed the survey (58%), 12 departments did not respond (36%), and 2 departments (6%) indicated they could not complete the survey. One of these departments explained their set shift structure included three weeks of a mix of daytime work, weekends, nights and days off. The other department used a varying system of days or nights worked per year, to be done at the discretion of the CCM physician, thereby precluding determination of the number of days off or shifts worked per week. Of the 19 departments that completed the data collection form, 44 total ICUs were covered, with 69 distinct CCM physician shifts. The mean time to complete the collection form per shift was 7.34 min (± 3.6 min), median 5 min. Overall, ten CCM physician shift types were identified and defined by their various combination of weekday and weekend work, described in (Table 1) and visualized in (Figure 2).

Of these 69 work shifts, there were 65 shifts that involved a CCM physician working with residents, fellows, and/or APPs, and 4 “CCM physician only” shifts (Table 2). The five most prevalent shifts noted were: 7 days in house with nighttime home call (n=18), 7 days in house without nighttime work (n=17), 7 nights in house without daytime work (n=10), 5 weekdays in house with weekend in house and home call (n=8), and 5 weekday nights in house without daytime or weekend work (n=5). ICU beds according to CCM physician work shifts can be found in (Table 3). We report quantitative data for shifts with at least five responses, which included five of the ten shift types. The number of beds distribution overall and by shift reveal a trend that when a CCM physician covers the ICU for 7 days in house with 7 nights home call, the number of beds appears to be less than the other shifts identified. Given the limited number of responses in this feasibility survey, we did not perform subgroup analysis.

Days off by work shift are categorized in Supplemental (digital content) by global setting (working with resident/fellow/APP or CCM physician only) and by shift type. This table illustrates how in a larger nationwide survey, the results could aid in documenting current trends in staffing patterns as days off according to CCM physician shift type. As in, we report quantitative data for shifts with at least five responses. Data reflected that for shifts involving 7 days of coverage (any combination of day or night), the average resulting days off ranged from 2.5 – 3.6 with a median of 3 days. Also reflected in this table is the percentage of respondents that received additional monetary compensation over their base pay for working the shift (% pay). All respondents reporting 5 nights of home call or in house every other night (total n = 5) received extra compensation. For the most common shifts reported, the percentage of respondents noting added compensation ranged from 38 to 80%. Quantifying the compensation was not asked. Supplemental (digital content) should not be considered as data for benchmarking, as that would require more than documenting current trends in staffing patterns.

Discussion

The proper number of days off following ICU coverage remains an enigma. Because of the apparent countless variations of CCM physician staffing models, collecting data via surveys has been difficult. Unlike other surveys, we created a data collection form that allowed us to analyze the data and subsequently categorize shifts, rather than predetermine shift categories. Our goal in this feasibility study was to determine if our novel form would result in meaningful data collection and allow for staffing model comparisons and defining of days off after a variety of CCM physician work shifts. Our results should not be extrapolated to national data. We do not include the geographical distribution of departments completing the survey, as we did not design the study to be randomized nor represent all geographical areas.

Time “off” from work can be thought of in multiple ways time to recover; a form of compensation; nonclinical time to develop professionally through academic pursuits such as research, regional or national society involvement, curricula development; or time off without clinical or administrative responsibilities allowing for one to seek personal bliss. Recovery from night in housework is different than night at homework or day work and as such requires different time allocations. Monetary compensation was partially addressed on the data collection form by asking if additional money was provided for working the ICU shift beyond their base pay. As the survey population consisted of anesthesiologist CCM physicians, another way to think of this question was to compare if they instead worked in the operating room for those shifts would their pay decrease (YES) or remain the same (NO). There is significant heterogeneity in CCM physician work shifts and the benefits of one staffing model over another have not been noted [5,6].

While there are no guidelines or a preferred staffing model, the Society of Critical Care Medicine (SCCM) and American College of CCM have commented that an ideal ICU model should have a dedicated CCM physician 24/7, especially for high acuity, high volume units [7]. However, implementation of 24/7 CCM staffing requires extensive institutional support. Regardless of staffing model, the presence of a CCM physician and multidisciplinary team has been found to reduce mortality rates and shorten ICU lengths of stay for a subset of units. Greater attention has been focused on the impact that CCM staffing models have on the CCM physician due to the growing recognition of burnout syndrome, especially within high acuity professions. In 2016, the SCCM created a task force that among other goals, was tasked to review issues related to critical care workload and burnout syndrome6. As predicted, ICU strain and increased staff workload, especially when patient to CCM physician ratio exceeds 15:1, may negatively impact wellness and is independently associated with symptom intensity.

Several studies have noted that volume and timing of clinical work, especially a heavy night shift burden or the number of consecutive days worked, are key factors that influence burnout likelihood. Future considerations for modifying staffing models, especially as a 24/7 model becomes more commonplace, should include strategies to reduce ICU workforce burden and decrease burnout rates. Our survey uniquely categorized results into several general work shifts after survey completion, due to collecting data specifically by hours worked and if the CCM physician was in house. Asking respondents to select from a list of pre-determined shifts was deemed less accurate since there is no standard ICU shift, and doing so might have led to fewer surveys being completed. We anticipate that after analyzing results of a national survey, there will be more work shift categories than the ten identified (e.g., tele ICU), although the ten defined are likely the most common.

Shaefi et al provides valuable data from their national clinical practice pattern survey of anesthesia CCM physicians. They report ~82% of respondents work in an academic practice (residents, fellows, APP), similar with our pilot data of 87%, with only ~2% in CCM physician only care as compared to our 6%. Most notably ~10% reported tele ICU coverage, which is not reflected in our data, nor is coverage at night with less consecutive days – like operating room call coverage. Like our data results, the most common nighttime coverage model was shifting based work within house coverage 24/7, followed closely by a nighttime home call after in house coverage model. Despite our limited and nonrandom data, the results appear like those found with a broader national survey.

Data from a national survey is anticipated to add value for leadership (e.g., department chairs, CCM division chiefs, and hospital/medical school executives) as it may help further elucidate an FTE for a CCM physician, aid in adjusting compensation to compete with the market of critical care staff and allow for appropriate time off. While benchmarking days off by shift type is an ideal goal, it presumes that current staffing patterns are appropriate for the work being done. In the proposed national survey, we aim to analyze for trends or any variation within a work shift. Additionally, with potentially more work shift categories to compare, there may be a stronger relationship between the number of beds and days off within a work shift category. At this time, no definitive conclusions should be made from the data collected. We believe this feasibility study demonstrates that our novel data collection tool allowed for expeditious data collection and results that may contribute towards analysis of staffing pattern trends of shift types and time off per work shift.

Our next step is working with the American Society of Anesthesiology’s CCM Committee to facilitate a national survey of anesthesiology departments. Starting data collection with academic departments is anticipated to be straightforward as chairs are readily identifiable. In addition, this approach may prevent data duplication by having division chiefs complete a survey for each unit staffed, rather than pursuing individual CCM physicians from CCM societies distribution lists.

References

1. Shaefi S, Pannu A, Mueller A, Brigid F, Adam E, et al. (2023) Nationwide Clinical Practice Patterns of Anesthesiology Critical Care Physicians: A Survey of Members of the Society of Critical Care Anesthesiologists. Anesth Analg 136(2): 295-307.
2. Haupt M, Bekes C, Brilli R, Linda C Carl, Anthony W Gray, et al. (2003) Guidelines on critical care services and personnel: Recommendations based on a system of categorization of three levels of care. Crit Care Med 31(11): 2677-2683.
3. Pastores S, Kvetan V, Coopersmith C, J Christopher F, Curtis S, et al. (2019) Workforce, workload, and burnout among intensivists and advanced practice providers: A narrative review. Crit Care Med 47(4): 550-557.
4. Siddiqui S, Warner M, Kelly L, Erika M, David JD, et al. (2023) Determinants of Professional Fulfillment and Burnout Among Intensivists: A National Survey by the Society of Critical Care Anesthesiologists in 2022. Anesth Analg 137(2): 375-382.
5. Arora R, Lee E, Kent D, Mina A, Yoan L, et al. (2021) Characterizing physician-staffing models in the care of postoperative cardiac surgical patients in Canada. CJC Open 3(11): 1365-1371.
6. Sabov M, Daniels C (2018) The value of 24/7 in-house ICU staffing 24/7 intensivist in the ICU. Crit Care Med 46(1): 149-151.
7. Pronovost P, Angus D, Dorman T, Karen AR, Tony TD, et al. (2002) Physician staffing patterns and clinical outcomes in critically ill patients: A systematic review. JAMA 288(17): 2151-2162.