Influenza is a major cause of morbidity and mortality.Reference Monto 1 – Reference Bresee, Reed and Kim 5 Hospitalized patients are at risk of exposure to influenza from other patients, visitors, and also from healthcare personnel (HCP) working while ill with influenza.Reference Horcajada, Pumarola and Martinez 6 – Reference Goins, Talbot and Talbot 10 To help prevent such exposures, the Centers for Disease Control and Prevention recommends that HCP should stay home until they are fever free for at least 24 hours after a febrile respiratory illness. 11 In addition, annual vaccination against influenza is recommended for all HCP. 11 Even though vaccination rates have increased in recent years, vaccination against influenza is still not universal among HCP.Reference Black 12 In addition, the effectiveness of the influenza vaccine varies considerably on an annual basis.Reference Flannery, Chung and Belongia 13 , Reference Flannery, Chung and Thaker 14 Thus, if HCP work with symptoms of influenza-like illness (ILI), influenza and other respiratory viruses might be transmitted to other HCP and patients under their care.
Presenteeism, working while ill, is common among HCP,Reference Chiu, Black and Yue 15 – Reference Jena, Baldwin, Daugherty, Meltzer and Arora 18 even among those working with immunocompromised patients.Reference Mossad, Deshpande, Schramm, Liu and Rothberg 19 Although individual reports in different settings indicate high levels of presenteeism, the scale of the problem, contributing factors, and the potential threat to patients is difficult to measure. Institutional policies to prevent HCP from working while ill with ILI are not standardized and are set at the discretion of each healthcare institution. This survey was undertaken to determine the range of institutional policies designed to prevent HCP from working with ILI.
Methods
The Infectious Diseases Society of America (IDSA) Emerging Infections Network (EIN) is a sentinel network of infectious diseases physicians who care for adult and/or pediatric patients in North America and who volunteer to participate in the network.Reference Pillai, Beekmann, Santibanez and Polgreen 20 In March 2018, the EIN conducted a mixed-methods survey of physician members who self-identify in the network as having hospital epidemiology responsibilities or interests (N = 728) to describe institutional experiences with and policies for HCP working with ILI. A 10-question survey was sent via an e-mailed web link or by facsimile, and an opt-out option was provided. Two reminders were sent to nonresponders at ~2-week intervals following initial survey. Members who had joined the EIN but had not yet responded to any surveys were excluded (N = 13), resulting in a denominator of 715 physician members.
The respondents were asked if their primary inpatient facility had occupational health policies regarding work restrictions for HCP with influenza or ILI. They were asked when the work-restriction policy was communicated to HCP and how it was enforced for different groups: attending physicians vs physicians in training, such as residents and fellows, vs nonphysician staff. Participants were also asked about policies regarding employee sick days and whether illness tracking occurred for patients and for HCPs (employed physicians and staff vs independent, nonemployed physicians) was ascertained. We also asked about the availability of influenza testing and treatment for employees and any provision of antiviral prophylaxis. The survey is provided in Appendix 1 (online). Data were analyzed using SAS version 9.4 software (SAS Institute, Cary, NC).
Respondents could also write comments for several of the questions or in a general comments section. For thematic analysis of the write-in comments, categoriesReference Morse 21 were developed through a data-driven process involving repeated examination of comments. Similar portions of text were identified and sorted so that commonalties could be determined. Labels were used to guide the sorting process.Reference DeCuir-Gunby, Marshall and McCulloch 22 A theme was a meaningful finding that emerged from the data.Reference Morse 21
Results
Of 715 infectious diseases physicians surveyed, 367 (51%) responded (Table 1). All regions of the United States were well represented. Respondents represented a full range of practices and hospital types, as well as years of experience in infectious diseases. Of 367 respondents, 135 (37%) were not aware of policies at their facility and opted out of the rest of the survey. Of the remaining 232 respondents, 206 (89%) reported that their inpatient facility had institutional policies regarding work restrictions for HCP with influenza or ILI. When asked when this policy was communicated, 145 (63%) answered that they received annual reminders either before and/or during influenza season, and 26 (11%) stated only that the policy was communicated on hire and the remainder (61, 26%) were not sure or did not answer the question.
Table 1. Practice Data for All 367 Respondents, Number (Column %), Infectious Diseases Society of America Emerging Infections Network, March 2018
Note: DOD, US Department of Defense.
The most commonly reported work-restriction enforcement strategies for nonphysician staff were sending HCP with ILI home and encouraging them to call in sick if necessary (reported by 119, 51%). Lack of enforcement of work restrictions was reported by 24% for physicians-in-training and by 42% for attending physicians but only by 13% for nonphysician staff (Fig. 1). Most respondents (124, 53%) reported that adherence to work restrictions was not being monitored, and 35 (15%) reported that adherence was monitored and 73 (32%) were not sure.
Fig. 1. Respondents’ perception of how work restriction policies are enforced for different types of HCP experiencing influenza-like illness. Infectious Diseases Society of America Emerging Infections Network, March 2018.
Respondents were provided a series of select-all-that-apply options to define their understanding of their facilities’ human resources policy for employee sick days. One hundred (43%) respondents reported that their facility employees had a defined number of sick days available per year, and 102 (44%) respondents reported that available days off for both vacation and illness were taken from the same pool. In addition, 45 respondents (19%) were unsure of their facility’s human resources policy for employee sick days, and 20 (9%) respondents reported an “other” policy. Also, 9 (4%) respondents also reported that additional sick days were made available during influenza season, with 3 stating that the leave was paid, 5 stating that it was unpaid, and 1 stating that pay status was not specified.
Laboratory-confirmed influenza is tracked by facilities most frequently for patients (223, 96% respondents), of which most (204 respondents) reported tracking all patients and a small number (19 respondents) reported tracking only nosocomial cases (Fig. 2). Smaller numbers of respondents reported tracking laboratory-confirmed influenza in employees (109, 47%), and only 13% (30 respondents) reported tracking influenza in independent physicians (those not employed by the facility).
Fig. 2. Does your facility track ILI or lab-confirmed influenza for patients, employees and independent physicians (not employed by the facility)? Infectious Diseases Society of America Emerging Infections Network, March 2018.
Influenza-like illness (ILI) in patients was tracked by the facilities of 85 (37%) respondents, with 58 tracking ILI for all patients and 27 tracking only inpatient nosocomial ILI. ILI in employees was tracked by the facilities of 53 (23%) respondents. ILI in independent physicians was tracked by the facilities of only 11 (5%) respondents. (Fig. 3a and 3b).
Fig. 3a. Does your facility provide influenza testing and/or treatment for employees? Infectious Diseases Society of America Emerging Infections Network, March 2018.
Fig. 3b. Does your facility provide antiviral prophylaxis after occupational exposure and/or on request after nonoccupational (eg, household exposure), Infectious Diseases Society of America Emerging Infections Network, March 2018.
Tracking of either ILI or lab-confirmed influenza was reported to be accomplished primarily by employee visits to occupational health (reported by 50 respondents in open-text field), using a call-in telephone line (reported by 29 respondents), and by laboratory/microbiology testing (reported by 23 respondents). In addition, 4 respondents reported use of an online or paper report form, and 9 stated that it is the manager or supervisor’s responsibility to report ILI to occupational health.
In another open-text field asking how tracking data were used, most respondents (n = 46) reported that the data were reviewed by or reported to either infection prevention or occupational health. A minority of respondents (n = 24) reported that an action was taken, including updating policies and practices, enforcing policies, offering prophylaxis or requiring wearing of masks by staff, and beginning investigations of potential clusters.
Overall, 63% reported that antiviral prophylaxis was provided to at least some employees after occupational exposures, although 9% reported provision of antiviral prophylaxis after nonoccupational (eg, household) exposures.
The survey contained several open-ended questions. Responses varied widely among the 169 respondents who described their institution’s occupational health policies regarding work restrictions. For example, some stated only that they have a policy (n = 45); others said the policy specified fever and a time period (usually 24 hours) to be afebrile off antipyretics (n = 79) before returning to work. For 12 respondents, a specific number of days exclusion were cited, (typically 5–7 days after illness onset), and 12 required approval from employee health to return to work. Interestingly, 20 reported combinations of the preceding criteria and other requirements. For example, 7 mentioned wearing a mask if respiratory symptoms were present, 3 specified additional restrictions for those working with immunocompromised patients such as in a NICU or transplant unit, and 6 described several time intervals (eg, total days of illness, days since illness onset, or days since resolution of fever without antipyretics) and stated that they would select whichever of these was longest.
Qualitative analysis of comments
Three other general themes were reflected among 51 respondents who provided additional comments. First, HCP had a perception that they were encouraged to work while ill. For example, one respondent noted, “Employees are penalized for using time off when ill even though it’s the policy that they stay home. … Comments from staff often speak to pressure (indirect or direct, internal or external) to come to work unless significantly ill.”
Second, several respondents described reasons for lack of compliance among physicians. “ ‘Presenteeism’ is most prevalent among doctors, for whom there is not a large pool of backups during the flu season, and many are uncomfortable asking others to fill in for them,” a physician noted, “And then there’s the inflated sense of self-importance.” Another commented, “It has been difficult with physicians or surgeons who do procedures and/or have busy outpatient clinics because of (the) need to cancel patients and/or inconveniences. There is no buffer in the system for another qualified provider to fill in.” Some described similarities in medicine and nursing, such as “Nurses and doctors work when sick. The reality is that if we didn’t—there would be very critical shortages of care givers during flu season. It’s unfortunate.”
The third theme related to potential loss of pay or vacation time related to policy structure (eg, having to use vacation time in lieu of sick leave). One respondent noted, “Lumping together ‘sick time’ and ‘vacation time’ is a significant contributing factor to presenteeism.” Others commented on sick leave impacting bonus structures because sick days are counted as days of work missed and noted that if sick days are unpaid, the financial element of not coming to work needs to be factored in. A respondent summarized it this way: “If they really wanted people to stay home when ill they would have dedicated sick time.”
Discussion
Our results demonstrate a wide range of reported facility-level policies for managing healthcare professionals with ILI. We found substantial differences in reported approaches for capturing presenteeism across hospitals and across different types of HCP. Most hospitals described by respondents do track influenza infection in patients, but they do not routinely track ill employees or independent physicians. In the absence of mechanisms to effectively track illness in HCP, it is difficult to enforce a policy specifying work restrictions. In general, illness, either laboratory-confirmed influenza or ILI, is tracked the least frequently for physicians, especially those not directly employed by the hospital. Some hospitals track only laboratory-confirmed influenza cases among HCP, whereas other hospitals were reported to track employees with ILI. Hospital policies are intended to prevent healthcare professionals from working with ILI. However, the reported enforcement of such policies was highly variable, and it may be limited by the frequently reported lack of effective tracking of illnesses. Finally, respondents reported substantial barriers, which may limit the ability of hospitals to monitor presenteeism. For example, many hospitals do not differentiate between sick days and vacation days, which limits the ability to determine whether absent healthcare workers are ill.
Presenteeism is not unique among healthcare professionals.Reference Kumar, Grefenstette, Galloway, Albert and Burke 23 – Reference Rousculp, Johnston, Palmer, Chu, Mahadevia and Nichol 26 The practice of professionals working while ill has been associated with lower levels of productivity and higher levels of errors and work-related injuries. Some reports estimate the cost of presenteeism, in general, to exceed the cost of absenteeism.Reference Whysall, Bowden and Hewitt 27 Presenteeism for people with ILI in the general workforce can also increase the spread of disease.Reference Kumar, Grefenstette, Galloway, Albert and Burke 23 , Reference Piper, Youk, James and Kumar 25 Influenza is a major cause of health-related absences.Reference Molinari, Ortega-Sanchez and Messonnier 28 , Reference Tsai, Zhou and Kim 29 In simulation studies, presenteeism substantially increases the number of influenza cases at a population level.Reference Kumar, Grefenstette, Galloway, Albert and Burke 23 Although similar studies are lacking in healthcare settings, the contribution of presenteeism may be substantial especially if influenza vaccination rates for healthcare professionals are low or if the efficacy of the annual influenza vaccination is low.
Causes of presenteeism noted in the literature include a sense of loyalty to colleagues, professional responsibility to patients, and feeling irreplaceable.Reference Chiu, Black and Yue 15 , Reference Szymczak, Smathers, Hoegg, Klieger, Coffin and Sammons 17 , Reference Giaever, Lohmann-Lafrenz and Lovseth 30 , Reference Chow and Mermel 31 Individuals who feel indispensable in their jobs are more likely to work while ill.Reference Bockerman and Laukkanen 32 Although we did not ask specifically about causes in this query, the comments support these points. Some evidence shows that vaccination against influenza may be associated with higher levels of presenteeism, possibly because vaccinated healthcare professionals may think that they are less likely to have influenza or to transmit it to their patients.Reference Chiu, Black and Yue 15 Other causes of presenteeism include fear of negative consequences from both administrators and colleagues, and, for many healthcare professionals, there may be financial consequences. Some of the respondents gave specific examples of financial consequences for missing work, for example, missing bonuses for absences regardless of the cause. More commonly, our results suggested that many institutions combine paid vacation and sick days. Thus, by claiming a sick day, a worker may lose a vacation day, creating a perverse incentive for presenteeism.
An effective approach to reducing presenteeism is having policies in place that support paid, dedicated sick leave,Reference DeRigne, Stoddard-Dare and Quinn 33 including for influenza.Reference Piper, Youk, James and Kumar 25 People with paid time off for illness are much less likely to work while ill. In contrast, the requirement for a “doctor’s note” for approved time off increases presenteeism.Reference Bockerman and Laukkanen 32 Combining sick days with vacation may seem more efficient in terms of human resource management practices, but it may result in putting patients and coworkers at risk when HCP come to work ill. Given that the influenza season is limited, it may be possible to implement strategies to minimize presenteeism during influenza season (eg, not exchanging sick days for vacation days during these months or providing additional ‘flu days’ during that time). Other approaches could include providing designated back-ups or alternative staffing mechanisms during the influenza season to help reduce the professional pressure for working while ill. Finally, enforcing existing or mandatory sick policies may be an important component of decreasing presenteeism. Further studies to test the impact of these interventions are needed. Having surveillance systems for detecting HCP illness and presenteeism in place may help to enforce current policies and measure existing compliance as well as to assess the effectiveness of interventions designed to reduce presenteeism.
Our study has several limitations. First, respondents had varying levels of information about presenteeism policies at their institution. Findings represent respondents’ perception or understanding of their institution’s policies. We did not survey healthcare administration or employee health, thus, a policy or practice (eg, monitoring) may have been in place of which the respondent was unaware. The EIN members were first asked about their awareness of presenteeism policies at their institutions; 63% (232 respondents) responded that they were aware of their facility policies and were asked to continue the survey. Thus, only the EIN members with specific knowledge about their institutions’ policies completed the survey. Second, the EIN is not a random sample of providers. Physicians who participate in the EIN may not necessarily represent the opinions of all infectious diseases physicians. In addition, EIN physicians who are specifically involved in policies or implementation and tracking of presenteeism might have been more likely to respond to this survey than others. However, the response rate for this survey (51%) was similar to other EIN queries, and we obtained responses from physicians in all sizes and types of hospitals and all US Census Bureau divisions. Third, this survey contained institutional-level questions, and some physician respondents may have represented the same institutions.
Despite the limitations, our results demonstrate a wide variation in the ability to measure HCP presenteeism with influenza and ILI and the policies designed to prevent it, as reported by survey respondents. Additional research to measure HCP illness and the frequency of working while ill is needed to ensure that facilities can best define progress in this important patient-safety issue. Such research should consider approaches to improve the ability to measure ILI among healthcare professionals that show up for work, which will help inform thoughtful and pragmatic policies designed to prevent presenteeism.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/ice.2019.305
Acknowledgments
The authors would like to acknowledge and thank Dr Marie A. de Perio and Dr Sophia Chiu, National Institute for Occupational Safety and Health, Centers for Disease Control and Prevention, for their review of and comments on this manuscript. The contents are solely the responsibility of the authors and do not necessarily represent the official views of the Centers for Disease Control and Prevention or the Department of Health and Human Services.
Financial support
This publication was supported by the CDC (cooperative agreement no. 1 U50 CK000477). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the Centers for Disease Control and Prevention or the Department of Health and Human Services.
Conflict of interest
P.M.P. reports travel and lodging support from 3M to present at a research meeting and personal fees from Eli Lilly for case adjudication for a clinical trial. All other authors report no conflicts.
