The research team, led by Zhaoping Liu and Bei Hu of the General Hospital of Central Theater Command of the People’s Liberation Army, together with colleagues from Hunan Normal University and other institutions, surveyed 700 nurses across three tertiary Grade A hospitals in Wuhan between October and December 2025. Tertiary Grade A hospitals represent the top tier of China’s hospital classification system, facilities where workloads are heavy, acuity is high, and the emotional demands on nursing staff are correspondingly intense. After excluding 20 invalid questionnaires, the researchers analyzed 680 valid responses, each completed anonymously and with informed consent under ethical approval from the Ethics Committee of Hunan Normal University.

Methodologically, the study leaned on a battery of validated instruments. Problematic short-video use was measured with the Problematic Short-Video Use Scale-9, a nine-item instrument designed to capture the loss of control, preoccupation, and functional impairment that characterize problematic use of platforms such as Douyin and TikTok. Perceived organizational support was assessed with the Perceived Organizational Support Scale, on-the-job functioning with the Stanford Presenteeism Scale, and the friction between work and home life with the five-item Work Interference with Personal Life subscale of the Work/Nonwork Interference and Enhancement Scale. Descriptive statistics and univariate comparisons were run in SPSS 27.0, while the centerpiece of the analysis, latent profile analysis, was conducted in Mplus 8.3, with multivariable multinomial logistic regression performed in R.

Latent profile analysis is a person-centered statistical technique, and its use here is what gives the study its distinctive texture. Rather than treating problematic short-video use as a single continuous score averaged across the sample, the method searches the data for subgroups of individuals who share similar symptom patterns, effectively asking whether the population contains natural clusters rather than one homogeneous blob. The answer was emphatically yes. Three distinct profiles emerged from the 680 nurses: a low-symptom use group of 383 nurses, or 56.3 percent of the sample; an escape-withdrawal prominent group of 226 nurses, or 33.2 percent; and a high-symptom comprehensive risk group of 71 nurses, or 10.5 percent.

The headline numbers are striking enough on their own. The mean total score for problematic short-video use across the sample was 21.23 with a standard deviation of 8.80, and 38.4 percent of the nurses were classified as at risk of problematic use. In other words, nearly two in five nurses at these flagship hospitals showed warning signs that their short-video habits had crossed from leisure into something potentially harmful. But the three-profile structure suggests that risk is not a single flavor. The largest cluster used short videos without notable symptoms. The middle cluster, more than a third of the sample, stood out for escape and withdrawal, patterns consistent with using the feeds to avoid negative emotions or disengage from stressors. The smallest but most concerning cluster, roughly one nurse in ten, scored high across the full spectrum of symptoms.

What pushes a nurse from one profile into another? The researchers found significant differences among the three latent classes in gender, department, perceived organizational support, work interference with personal life, and presenteeism, with P values below 0.05. These variables sketch a picture in which the workplace itself is woven into the pattern of screen use. Nurses who felt less supported by their organizations, who reported that their jobs bled more heavily into their personal lives, and who struggled to function while at work were disproportionately represented in the riskier profiles. Department mattered too, hinting that the specific texture of a nursing unit, from emergency medicine to respiratory wards, shapes the stress-recovery balance that short videos appear to fill.

The regression analysis sharpened the story further. After controlling for gender, department, perceived organizational support, and work interference with personal life, presenteeism remained significantly associated with membership in both the escape-withdrawal prominent group, with an odds ratio of 1.392 (95 percent confidence interval 1.115 to 1.738, P = 0.003), and the high-symptom comprehensive risk group, with an odds ratio of 2.365 (95 percent confidence interval 1.641 to 3.409, P < 0.001). An odds ratio above one means that higher presenteeism, the phenomenon of physically showing up for work while functioning below par, increases the odds of belonging to the riskier profiles. The effect was more than twice as large for the comprehensive risk group as for the escape-withdrawal group, suggesting a dose-response-like gradient in which the most impaired nurses at work were also the most symptomatic users of short videos.

Presenteeism is a concept that has quietly become one of the most consequential ideas in occupational health. Unlike absenteeism, which removes a worker from the schedule and is therefore visible in staffing rosters, presenteeism keeps the impaired worker on the floor, performing at reduced capacity. In nursing, where medication checks, dose calculations, and continuous monitoring leave little margin for error, the costs of reduced on-the-job effectiveness can be measured in patient safety. The new findings suggest a possible feedback loop: nurses whose recovery is insufficient may function poorly at work, and those same nurses may be the ones retreating into compulsive short-video use during their off-hours, further crowding out the restorative activities, sleep, exercise, and social contact, that would rebuild their capacity.

The study’s authors frame the results as a call for stratified, profile-informed management rather than one-size-fits-all interventions. Because the three latent classes differ in their dominant symptom patterns, they argue, nursing managers could use latent class characteristics to identify which nurses need which kind of support. The escape-withdrawal prominent group, for instance, may benefit most from interventions targeting emotional avoidance and stress coping, while the high-symptom comprehensive risk group may require more comprehensive attention to occupational functioning and recovery resources. The researchers specifically recommend incorporating presenteeism, departmental context, and recovery resources into occupational health management for nurses, an approach that treats screen behavior not as an isolated personal failing but as a signal embedded in the ecology of the workplace.

As with any cross-sectional study, the design has limits that the findings cannot escape. The data capture a single moment in time, so the direction of the arrow between presenteeism and problematic short-video use cannot be established with certainty; exhausted nurses may scroll to cope with impaired workdays, or compulsive scrolling may degrade the sleep and recovery that would have made those workdays less impaired, and both mechanisms could operate simultaneously. The convenience sampling of 680 nurses from three Wuhan hospitals also means the sample may not generalize to all nursing workforces, in China or elsewhere. Still, the scale of the risk identified, with 38.4 percent of nurses at risk of problematic use and one in ten in the highest-risk profile, gives the findings real weight for hospital administrators everywhere. Short-video platforms are engineered to capture attention with variable-reward loops that exploit the same neural circuitry as other compulsive behaviors, and workers in chronically stressed, recovery-starved professions are precisely the audiences most vulnerable to that design. The Wuhan study does not claim that short videos are uniquely dangerous among digital distractions, but it does something arguably more useful: it maps, with statistical precision, who is being pulled in, how their risk is structured, and where the leverage points for intervention may lie. For a profession already stretched thin by staffing shortages and burnout, the finding that nearly four in ten nurses show signs of problematic use of the very technology they reach for to unwind is a data point that hospital leaders, platform designers, and occupational health researchers cannot afford to scroll past.

Subject of Research: Problematic short-video use and presenteeism among clinical nurses in tertiary hospitals

Article Title: Latent profiles and associated factors of problematic short-video use among nurses in tertiary hospitals: a cross-sectional study

Article References: Liu, Z., Li, M., Liu, Y., Yang, P., He, X., Liu, W., & Hu, B. (2026). Latent profiles and associated factors of problematic short-video use among nurses in tertiary hospitals: a cross-sectional study. BMC Nursing. https://doi.org/10.1186/s12912-026-05451-9

Image Credits: AI Generated

DOI: 10.1186/s12912-026-05451-9

Keywords: nurses, short-video use, presenteeism, latent profile analysis, occupational health, burnout, perceived organizational support, work-life interference, BMC Nursing, cross-sectional study, digital addiction, Wuhan