On a fictional emergency shift, a resident wakes up with a low fever, myalgia and a dry cough. He hesitates: if he misses work, a colleague inherits the overload; if he goes, he may spread the virus to frail patients and to an already short-staffed team. He decides to go "only until noon". At 11 he coughs at the station and sees an elderly woman with COPD. No one takes his temperature. The case is not rare: the literature calls this sickness presenteeism, and the cost does not show up on the time clock.

This text gathers what studies and reviews have measured about physicians and other health professionals who work while sick: prevalence, reasons, risk of contagion and what training reinforces. In each section, the text reports the design and sample size and separates what was measured from what is our inference. The neighboring topic of sleep deprivation on shift deals with fatigue; here the axis is active illness and the decision to stay or leave.

Key points

• In Juiz de Fora, 82.8 percent of 309 physicians reported having practiced presenteeism; 57.6 percent fell ill at work and 35 percent took leave (Oliveira et al., 2022).
• In training in the US, 60 percent felt obligated to work while sick and 33 percent with flu-like symptoms; residents and students more than faculty (Kaldjian et al., 2019).
• Among 255 professionals with symptomatic COVID-19, 49.8 percent worked with symptoms (Linsenmeyer et al., 2023).
• The systematic review by Lui and colleagues retained 24 studies and concluded that the heterogeneity does not allow conclusive evidence on exposures, costs or interventions (Lui et al., 2018).
• Recurring reasons: not overloading colleagues, a sense of duty and "mild" symptoms; the hidden culture of the shift reinforces going in (Kaldjian et al., 2019; Widera et al., 2010).

• What sickness presenteeism is
• What Brazilian physicians reported
• Hidden culture in training
• COVID-19: presenteeism measured day to day
• What the reviews can (and cannot) state
• Five signs to stay home (our synthesis)
• What changes in shift reasoning
• How NEXO helps train this reasoning
• Frequently asked questions
• Take-home messages

What sickness presenteeism is

Presenteeism, in this article, is working despite an illness that reduces performance or raises risk to others. It is not to be confused with overtime or with "being present" in the sense of engagement. The distinction matters because a sick physician can transmit infection, make errors under fever and delay his own recovery (Widera et al., 2010).

Sickness presenteeism versus productivity presenteeism

Part of the organizational literature measures lost productivity with scales, without requiring a contagious infection. Lui's review separated exposures and outcomes in hospital physicians and nurses and showed that definitions vary between studies (Lui et al., 2018). For the bedside, the practical question is different: is there fever, vomiting, diarrhea or respiratory symptoms that should keep the professional away today?

Why the topic is about patient safety

Outbreaks linked to sick professionals are documented for influenza, norovirus and other agents (Widera et al., 2010). Presenteeism stops being only about the physician's well-being and becomes a patient safety item, in the same spirit as failures in the discharge transition: not transferring hidden risk to someone who did not consent.

One synthesis of ours is that professional duty includes protecting the patient from the professional himself when the illness is contagious or disabling.

What the classic numbers showed

Before the COVID-19 cohorts, narrative reviews already painted high prevalences. Widera and colleagues gathered examples that still circulate in hospital infection discussions (Widera et al., 2010).

Norwegian physicians and residents with flu

Rosvold and colleagues surveyed 1,015 Norwegian physicians: in one year, 80 percent went to work with an illness for which they would send a patient home (Widera et al., 2010). In a study on influenza vaccination, 70 percent of residents and students admitted having worked with flu-like symptoms (Widera et al., 2010).

Pediatric hospital in Philadelphia

Szymczak and colleagues applied an anonymous survey in 2014 at a large pediatric hospital in Philadelphia and received responses from 280 attending physicians and 256 advanced practice professionals (Szymczak et al., 2015). Although 95.3 percent believed that working while sick put patients at risk, 83.1 percent worked while sick at least once in the previous year (Szymczak et al., 2015). Respondents would go to work with significant acute respiratory symptoms (55.6 percent) and even with fever (16.0 percent), and physicians reported this more than the other professionals (Szymczak et al., 2015). Among the reasons, 98.7 percent cited not wanting to leave colleagues in the lurch, and 64.9 percent of the free-text comments described extreme difficulty finding coverage (Szymczak et al., 2015).

Influenza-like illness in a national survey

Chiu and colleagues, from the CDC, used a non-probability panel of 1,914 health professionals in the 2014-2015 influenza season (Chiu et al., 2017). Influenza-like illness was reported by 414 (21.6 percent), and 41.4 percent of them worked while sick, with a median of 3 days (Chiu et al., 2017). Pharmacists (67.2 percent) and physicians (63.2 percent) had the highest frequencies, and the hospital was the setting with the most working while sick (49.3 percent) (Chiu et al., 2017).

Outbreak in a long-term care facility

In the investigation of acute gastroenteritis in a nursing home, 94 percent of sick employees went to work and 8 percent vomited at work (Widera et al., 2010). The case illustrates the extreme: clear symptoms and still showing up.

A cautious reading

These numbers come from different contexts and instruments. They serve as a warning, not as a meta-analysis. The later systematic review by Lui reinforced that the quality of the measures limits causal conclusions (Lui et al., 2018). Anyone who reads only the percentage without the exact question is comparing different things.

What Brazilian physicians reported

Oliveira and colleagues interviewed 309 physicians in Juiz de Fora, Minas Gerais, with a questionnaire and the WHOQOL-BREF (Oliveira et al., 2022). The sample was 54.4 percent women, with a mean age of 40.9 years (Oliveira et al., 2022).

Illness, sick leave and presenteeism

Among the participants, 57.6 percent (n = 178) reported having fallen ill during professional activity, 35 percent (n = 110) took sick leave and 82.8 percent (n = 256) had already practiced presenteeism (Oliveira et al., 2022). The most cited illnesses involved the respiratory system (29.5 percent), infectious and parasitic diseases (14.38 percent) and the circulatory system (9.59 percent) (Oliveira et al., 2022).

Quality of life and time in the profession

In the multivariate analysis, the environment domain had the lowest mean on the WHOQOL-BREF, and 10 years or more in the profession was associated with a higher chance of a higher score in that domain, with OR 2.28 (95 percent CI 1.46 to 3.56) (Oliveira et al., 2022). Age 39 years or older had a protective effect in the same model (Oliveira et al., 2022). About 80.9 percent did not feel represented by organizations regarding quality and safety at work (Oliveira et al., 2022).

What this changes in Brazilian practice

The combination of high presenteeism and relatively low sick leave suggests a cultural norm, not just an individual failure. It connects with burnout during training and with moral injury: the physician may go to work sick out of a distorted sense of duty, not out of disregard for transmission.

The Brazilian literature from Oliveira also reminds us that respiratory diseases lead the self-reported illness at work (Oliveira et al., 2022). This brings presenteeism close to the calendar of respiratory viruses and to seasonal leave policy. Services that discuss the topic only during a pandemic forget the following winter.

Hidden culture in training

Kaldjian and colleagues surveyed students, residents and faculty in Family Medicine, Pediatrics and Internal Medicine in 2012-2013 about working with a possible infectious illness (Kaldjian et al., 2019). In the second survey, 127 people responded (44 percent response rate): 69 students, 20 residents and 38 faculty (Kaldjian et al., 2019).

Obligation and flu-like symptoms

Sixty percent reported a professional obligation to work while sick. Thirty-three percent felt obliged to work with flu-like symptoms (fever, myalgia and cough), with a difference between residents (67 percent), students (35 percent) and attending physicians (14 percent; p = 0.001) (Kaldjian et al., 2019).

Stated reasons

The majority (83 percent) wanted to avoid creating more work for colleagues. Residents and students also feared negative repercussions and seeming lazy more often than attending physicians (Kaldjian et al., 2019). The authors discuss a hidden curriculum of efficiency that competes with the safety culture (Kaldjian et al., 2019).

This finding connects to the post about the hidden curriculum in Medicine: what is learned in the hallway can cancel out what is taught about isolation and staying away.

Design limitations

The 44 percent response rate opens up selection bias. The survey is from a single center and a decade ago. Even so, the resident-attending gradient is a useful signal for preceptorship: whoever teaches models the decision to go in or stay home.

COVID-19: presenteeism measured day by day

Linsenmeyer and colleagues analyzed occupational health interviews and a follow-up questionnaire in the veterans health system in Boston, USA (Linsenmeyer et al., 2023). Among professionals with COVID-19, 255 (78.0 percent) were symptomatic; of these, 127 (49.8 percent) reported presenteeism at the time of diagnosis (Linsenmeyer et al., 2023).

Specific and nonspecific symptoms

Among 168 with at least one typical COVID-19 symptom, 79 (47 percent) worked while symptomatic; among 87 with only nonspecific symptoms, 47 (54 percent) worked (Linsenmeyer et al., 2023). The difference between these groups was not significant (Linsenmeyer et al., 2023).

Why they went to work

In the subgroup that answered the follow-up (52 people, 20.4 percent), concerns about colleagues' workload (66 percent) and personal responsibility (45 percent) outweighed paid leave limits (19 percent) and the perceived expectation to work while sick (10 percent) (Linsenmeyer et al., 2023). Symptoms were attributed to allergy, a cold, migraine or "mild COVID" (Linsenmeyer et al., 2023).

Japanese physicians on the front line

Ishimaru and colleagues analyzed 3,968 employed physicians in a national survey (response rate 18.5 percent) during the pandemic (Ishimaru et al., 2023). Here presenteeism was measured as impaired functioning at work, with 21 points or more on the Work Functioning Impairment Scale, and occurred in 13.9 percent (Ishimaru et al., 2023). Front-line physicians had aOR 1.71 (95 percent CI 1.16 to 2.53) and second-line physicians 1.45 (1.17 to 1.78) compared with the others (Ishimaru et al., 2023). Involvement in COVID-19 vaccination had adjusted OR 0.94 (0.77 to 1.16), an interval that includes the null and is inconclusive for association (Ishimaru et al., 2023).

The 13.9 percent rate (Ishimaru et al., 2023) is lower than that of Oliveira and of Linsenmeyer because the question (functional impairment, not going to work sick), the time window and the population differ: comparing crude prevalences between studies without aligning the definition leads to misreading.

What the reviews can (and cannot) state

Lui and colleagues identified publications on presenteeism in hospital physicians and nurses, selected 38 for quality and retained 24 for extraction (Lui et al., 2018). Seventeen addressed exposures and outcomes, four costs, one intervention and two validations (Lui et al., 2018).

Heterogeneity and a null conclusion

The explicit conclusion: the heterogeneity and limited quality of the tools did not produce conclusive evidence on the association of presenteeism with exposures, economic costs or interventions (Lui et al., 2018). This does not say the problem is small; it says the available science does not support a single measurable effect.

Review on infectious disease

Webster and colleagues systematically reviewed presenteeism linked to infectious disease: of 3,580 articles, they included 23 articles with 24 studies, almost all cross-sectional and of relatively low quality (Webster et al., 2019). Prevalence ranged from 35 to 97 percent, and the most consistent risk factor was occupation: working in health care, and specifically being a physician, raised the risk (Webster et al., 2019). The reasons grouped into organizational policy and culture, job characteristics such as lack of coverage, and personal reasons such as not overburdening colleagues (Webster et al., 2019).

Recent scoping review

Sampaio and colleagues, in a scoping review published in 2026, included 31 articles, mostly cross-sectional (70.97 percent), with prevalences ranging from 14.1 percent in Australia to 100 percent in Iran, depending on the instrument and the population (Sampaio et al., 2026). Among 90 mentions of reasons, the most frequent were mild symptoms (36, or 40.0 percent) and difficulty finding a replacement (25, or 27.8 percent) (Sampaio et al., 2026).

How to read the prevalence table

Source	Population	Main finding	Reading note
Oliveira 2022	309 physicians (Brazil)	82.8 percent had already practiced presenteeism (Oliveira et al., 2022)	Cumulative self-report
Kaldjian 2019	127 in training (USA)	60 percent obligation; 33 percent flu-like (Kaldjian et al., 2019)	Attitude, not just behavior
Linsenmeyer 2023	255 HCW with symptomatic COVID	49.8 percent worked while symptomatic (Linsenmeyer et al., 2023)	Counts any symptomatic period
Ishimaru 2023	3,968 physicians (Japan)	13.9 percent; aOR versus 1.71 (1.16 to 2.53) (Ishimaru et al., 2023)	Functional impairment, not going in sick
Chiu 2017	414 professionals with influenza-like illness (USA)	41.4 percent worked while sick; physicians 63.2 percent (Chiu et al., 2017)	Online panel, not probabilistic
Lui 2018	24 studies retained	Inconclusive evidence (Lui et al., 2018)	Heterogeneity of measures

The table does not rank countries: it ranks different questions. The clinician uses the numbers to calibrate local risk, in the spirit of confidence calibration.

Five signs to stay home (our synthesis)

A validated universal score for "can the physician work today?" is missing. The list below is our synthesis based on the infection patterns and the reasons cited in the studies, not a published score.

1. Measured fever or a clear feeling of fever in the last 24 hours.
2. Vomiting, acute diarrhea or inability to hydrate during the shift.
3. New cough with systemic malaise during a respiratory transmission period.
4. Infected skin lesions draining in an area that protective equipment does not cover.
5. Cognitive impairment or decision fatigue intense enough to doubt one's own judgment.

What the team can agree on before the crisis

Services that define last-minute coverage, an alternative teleconsultation and absence without informal punishment reduce the incentive to go in sick (Widera et al., 2010; Linsenmeyer et al., 2023). Our hypothesis: the resident can only stay home if the preceptor models the same rule and if the schedule has a real replacement.

Quick checklist before entering the hospital (our synthesis):

• Temperature and symptoms in the last 24 hours.
• Unit with high-risk patients today.
• Name of the replacement or coverage path.
• If the answer to "would I keep the patient away with what I feel?" is yes, the bar to go in rises.

What changes in on-call reasoning

Presenteeism affects clinical reasoning in two ways: transmission and performance. The first is epidemiological; the second resembles the logic of sleep deprivation, but the cause is illness, not just wakefulness.

Contagion and differential diagnosis

A professional with influenza or norovirus becomes an intramural vector (Widera et al., 2010). Vulnerable patients, such as older adults with delirium or the immunosuppressed, pay the price. The decision to stay away is a clinical act, not just an administrative one. During influenza periods, the question gains extra weight.

When the error fits in the conversation

If the physician worked while ill and harm occurred, honest communication follows the same ethics as disclosing a diagnostic error: facts, responsibility and plan, without dramatizing blame.

Safety net for the team

Agreeing on warning signs and who covers the absence is a safety net applied to the team itself: the indirect patient of presenteeism is the one in the next bed.

In shift practice, presenteeism competes with other pressures already discussed in NEXO: alert fatigue in the digital hospital, the rush of the patient schedule and the culture of not interrupting the flow. None of these pressures justifies exposing a neonate, an oncology patient or a frail elderly person to a virus brought by the on-call physician. The useful question at handoff is objective: is anyone on the team sick enough to keep away a patient with the same condition? If the answer is yes, the roster needs a plan B in the same shift, not the next day.

Another practical point is documenting the leave without stigma. Services that record "leave due to infectious disease" as an expected event, not a moral failure, align personnel records with patient safety. Our synthesis: the silent heroism of the sick person who shows up is a negative culture indicator, not a resume praise.

Role of nursing and the roster

Presenteeism is not only a medical decision. Nurses and technicians appear in several of the studies gathered by Lui and by the scoping review, with prevalences that sometimes exceed those of physicians depending on the instrument (Lui et al., 2018; Sampaio et al., 2026). A multiprofessional roster that has no coverage for one link breaks the next link: the resident who stays home without a substitute pressures the on-call physician; the sick on-call physician pressures nursing.

What nursing sees first

Early signs of an impaired colleague (persistent cough, vomiting in the bathroom, atypical slowness) are often noticed by those who share the station. Clear leave policies protect those who report the risk without becoming a "snitch". This aligns with the logic of clinical deterioration and escalation: the alert channel must exist before the crisis.

Coverage as safety technology

When there is no backup list, actionable hour bank or planned redistribution, the "stay home" discourse is empty. Our hypothesis: the most cost-effective intervention is not a poster about hygiene, but a coverage protocol that the roster chief can execute in 30 minutes.

The final message for the clinical reader is sober: presenteeism due to illness is frequent, measurable and modifiable when the institution treats leave as a protection tool, not a character flaw. As long as the informal metric is "who came even though sick", the numbers from Rosvold, Oliveira and the COVID-19 studies will repeat with other viruses and other teams.

Three scenes to discuss in preceptorship

Three clinical scenes help fix the reasoning (all fictional, but aligned with the mechanisms of the studies):

1. Resident with measured fever and cough at the start of the neonatal ICU shift: the risk of transmission outweighs any argument of "short staff".
2. Surgeon with acute gastroenteritis who insists on operating electively: the parallel with the nursing home outbreak in which 94 percent went to work sick shows the pattern, not the exception (Widera et al., 2010).
3. Preceptor who praises the "warrior" intern who came with the flu: this reinforcement is exactly the hidden curriculum measured by Kaldjian (Kaldjian et al., 2019).

In each scene, the correct question is not "can you handle it?". It is "who pays the risk if you stay?". When the risk payer is the patient, professional ethics points to leave with coverage, not heroism.

From a teaching standpoint, it is worth asking the resident to write, once a month, a paragraph about the last time they almost went to work sick and what made them change their mind (or what made them go). This cheap exercise makes the local norm visible. If most reports end with "I went because no one was covering", the problem is not motivational: it is about the roster. If they end with "I went because the boss would praise me", the problem is hidden curriculum and evaluation. The data from Kaldjian and Linsenmeyer point to both mechanisms at the same time (Kaldjian et al., 2019; Linsenmeyer et al., 2023).

Reading errors this post avoids

Three recurring confusions hinder the clinical conversation.

1. Treating any percentage of presenteeism as comparable across countries without aligning the question.
2. Concluding that "no evidence" in Lui's review means "no harm".
3. Using presenteeism as a synonym for burnout or for sleep deprivation.

Each confusion changes the approach: the first generates a sterile ranking debate; the second paralyzes local policy; the third mixes interventions. The safe path is to cite the study with n, definition and CI when available, and to mark our synthesis when the bridge is an extrapolation.

When the CI includes the null

In the Japanese survey, the association between working in COVID-19 vaccination and presenteeism had OR 0.94 with 95 percent CI from 0.77 to 1.16 (Ishimaru et al., 2023). Because the interval crosses 1, the result is inconclusive for benefit or harm of this exposure on the measured outcome. Reporting only the OR point without the CI would invite a false reading of "protection".

How NEXO helps train this reasoning

In NEXO, the case of the day is fictional, with a real clinical presentation and up to six progressive clues, and you propose the diagnosis by searching for the ICD-10 code. One clue may bring "on-call physician with fever and cough attended older adults the day before"; another, "three patients from the same ward with flu-like symptoms within 48 hours". The five-level proximity feedback helps separate influenza, COVID-19, another respiratory virus and bacterial pneumonia, and reminds you that the source of the outbreak may be the team itself.

Frequently asked questions

Is physician presenteeism the same as burnout?

Burnout is an exhaustion syndrome; presenteeism due to illness is working despite an illness. They are not the same entity. They can coexist, but the measures and interventions differ (Oliveira et al., 2022; Lui et al., 2018).

Does a cold with only a runny nose justify working the shift?

It depends on the context, on fever, on the unit (oncology, neonatal, immunocompromised patients) and on local policy. Studies show that mild symptoms are a frequent reason to go in (Sampaio et al., 2026). When in doubt with risk to vulnerable people, the safest option is to stay away and arrange cover.

Does Lui's review prove that presenteeism causes no harm?

It shows an absence of conclusive evidence due to heterogeneity of measures, not evidence of absence of harm (Lui et al., 2018).

Why do residents go in sicker than attending physicians?

In Kaldjian's study, residents reported more obligation with flu-like symptoms and more fear of informal punishment (Kaldjian et al., 2019). Hierarchy and evaluation shape behavior.

What can the service do this week?

Define explicit criteria for staying away, guarantee actionable cover and align preceptors so they do not praise those who came in sick. Without real cover, the safety discourse becomes a hidden curriculum in reverse.

Take-home messages

• Presenteeism due to illness is common among physicians and other health professionals; prevalences vary with the question and the population (Webster et al., 2019; Oliveira et al., 2022).
• Central reasons: not overburdening colleagues and a sense of duty, including with flu-like symptoms (Kaldjian et al., 2019; Szymczak et al., 2015).
• In COVID-19, almost half of the symptomatic professionals in a Boston health system worked with symptoms (Linsenmeyer et al., 2023).
• Systematic reviews still do not settle on a single causal association; heterogeneity leaves findings inconclusive on several questions (Lui et al., 2018).
• The decision to stay home is an act of patient safety and needs institutional cover, not only individual heroism (Widera et al., 2010).
• Train the contagion question together with the cognitive performance question at the start of every shift (our synthesis).

References

1. Widera E, Chang A, Chen HL. Presenteeism: a public health hazard. J Gen Intern Med. 2010;25(11):1244-1247. PMID: 20549378. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2947637/
2. Szymczak JE, Smathers S, Hoegg C, et al. Reasons why physicians and advanced practice clinicians work while sick: a mixed-methods analysis. JAMA Pediatr. 2015;169(9):815-821. PMID: 26146908. https://pubmed.ncbi.nlm.nih.gov/26146908/
3. Chiu S, Black CL, Yue X, et al. Working with influenza-like illness: presenteeism among US health care personnel during the 2014-2015 influenza season. Am J Infect Control. 2017;45(11):1254-1258. PMID: 28526310. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5670002/
4. Lui JNM, Andres EB, Johnston JM. Presenteeism exposures and outcomes amongst hospital doctors and nurses: a systematic review. BMC Health Serv Res. 2018;18(1):985. PMID: 30567547. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6299953/
5. Kaldjian LC, Shinkunas LA, Reisinger HS, et al. Attitudes about sickness presenteeism in medical training: is there a hidden curriculum? Antimicrob Resist Infect Control. 2019;8:149. PMID: 31508227. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6727412/
6. Webster RK, Liu R, Karimullina K, et al. A systematic review of infectious illness presenteeism: prevalence, reasons and risk factors. BMC Public Health. 2019;19(1):799. PMID: 31226966. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6588911/
7. de Oliveira JM, Gonçalves LSB, da Fonseca ALS, et al. Physicians' quality of life, illness and presenteeism: a cross-sectional epidemiological study. Rev Bras Med Trab. 2022;20(3):401-411. PMID: 36793455. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9904841/
8. Linsenmeyer K, Mohr D, Gupta K, et al. Sickness presenteeism in healthcare workers during the coronavirus disease 2019 (COVID-19) pandemic: an observational cohort study. Infect Control Hosp Epidemiol. 2023;44(10):1693-1696. PMID: 37039605. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10587373/
9. Ishimaru T, Yoshikawa T, Okawara M, et al. Presenteeism in front-line physicians involved in COVID-19-related clinical practice: a national survey of employed physicians in Japan. Environ Health Prev Med. 2023;28:13. PMID: 36740269. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9922563/
10. Sampaio CL, Soares TC, Grangense VLC, et al. Presenteeism among healthcare professionals in a hospital setting: a scoping review. Rev Lat Am Enfermagem. 2026;34:e4900. PMID: 42585480. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13459727/