A Retrospective Analysis to Evaluate Seasonal Pressure Injury Incidence Differences Among Hip Fracture Patients in a Tertiary Hospital in East China

Patients with a hip fracture are at high risk for pressure injury. A retrospective review of the electronic medical records of consecutive patients with a hip fracture treated in 2016 in a single tertiary hospital in east China were examined to investigate whether time of year affected the incidence of hospital-acquired pressure injury.Data collected included demographic characteristics (patient name, hospital number, age, gender, and body mass index); possible risk factors for pressure injury, such as presence of diabetes mellitus, admission hemoglobin, admission albumin, length of surgery, and the lowest and/or last Braden Scale score before pressure injury developed; and pressure injury information, which included time of occurrence (days after surgery), location (sacrum and coccyx, ischial tuberosity, or heel), stage, and treatment outcome. Pressure injury incidence was calculated with 95% confidence intervals (CIs) in each month and season (spring, summer, autumn, and winter). Odds ratios (ORs) and 95% CIs were calculated as estimates of risk. Multivariate logistic regression was used for risk factors. Of the 235 patients with a hip fracture included in the study, 95 (40.4%) were male, 140 (59.6%) were female, and mean age was 70.4 ± 10.5 (range 48–81) years. Thirty-one (31) patients (13.2%, 95% CI 9.1%-18.2%) developed 37 pressure injuries, 30 of which (81.1%) were Stage 1. The incidence of pressure injury was lowest in November (5.0%; 95% CI: 0.0%-24.9%) and highest in June (22.7%; 95% CI: 7.8%-45.4%). Average Braden Scale scores (14.2 ± 3.2) were lower in June than in November (16.6 ± 3.5), owing to differences in the Braden Scale skin moisture subscale. Seasonally, the incidence of pressure ulcers was 20.8% (95% CI: 12.2%-32.0%) in the summer compared to 7.5% (2.5%-16.6%) in autumn (crude OR 3.3; 95% CI:1.0-12.1; P = .025). Multivariate logistic regression analysis showed the Braden Scale score was the only independent risk factor (P <.05) for pressure injury incidence. Adjusting for the Braden score, the OR of summer season was 1.537 (95% CI: 0.964-2.872). The findings suggest that humidity and temperature levels, which are very high in the summer in China, may affect pressure injury incidence and that the Braden Scale score — especially the skin-moisture level subscore — is a valid predictor of pressure injury risk in this population. While more research is needed, additional pressure injury prevention strategies should be provided for patients hospitalized with a hip fracture in the summer months.

Hip fractures are a common trauma in older adults and as the number of elderly persons increases, the incidence of hip fractures likely will increase as well.^1-3 Following surgical repair, these patients are at high risk for pressure injury.⁴ Previous studies^4-7 in these populations have reported a pressure ulcer incidence of 3.4% to 59.8%. Pressure ulcers in patients with hip fractures have been shown in retrospective analyses to lengthen hospital stay⁸ and increase health care costs.⁹ Moreover, pressure ulcers have been shown in a retrospective study¹⁰ (N = 343) to decrease patient survival after a hip fracture; risk for reduced survival to hospital discharge is markedly increased (hazard ratio [HR] 4.25, 95% confidence interval [CI]: 1.35–13.36; P = .013) as is risk for reduced survival at 1 year after surgery (HR 4.15; 95% CI: 2.14–8.06; P <.001), making it crucial to prevent this complication in patients with a hip fracture. 

The Braden Scale is an important tool for assessing patient pressure ulcer risk and has been shown in a prospective study¹¹ and a systematic review¹² to offer the best balance between sensitivity and specificity and the best risk estimate. Several studies,^7,13-15 including a prevalence and incidence study,⁷ a secondary analysis of data,¹³ and a prospective cohort study,¹⁵ have investigated risk factors for pressure injury among hip fracture patients. This research found factors such as older age, dehydration, longer wait before surgery, longer intensive care unit stay, longer surgical procedure time, cognitive impairment, higher American Society of Anesthesiologists grade, and impaired mobility were possible risk factors for pressure ulcers in this population. 

In the authors’ clinical practice, the incidence of pressure injury in hip fracture patients was observed anecdotally to increase in the summer. Literature about the possible effect of seasonal changes on the incidence of pressure injury in hip fracture patients in east China could not be identified. The purpose of this retrospective study was to examine if the incidence of pressure injury in hip fracture patients varies by month or season. 

Patient population. Discharge data from the electronic medical records of a single tertiary hospital in China were collected and abstracted. Inclusion criteria stipulated participants should be consecutive patients with a hip fracture treated between January 2016 and December 2016 whose records indicated risk for pressure injury pressure ulcer as assessed using the Braden Scale, as well as the presence of a pressure injury Patients with a hip fracture whose pressure injury developed before admission or who died during hospitalization because of serious complications or whose pressure injury records were missing were excluded from the study.

Ethical consideration. Data collection was approved by the medical ethics committee of the Affiliated Hospital of Nantong University for anonymous patient inclusion.

Data collection. The data collected included demographic characteristics (patient name, hospital number, age, gender, and body mass index [BMI]); possible risk factors for pressure injury, such as diabetes mellitus, admission hemoglobin, admission albumin, length of surgery, and the lowest and/or last Braden Scale score before pressure injury developed; pressure injury information, which included time of occurrence (days after surgery), location (sacrum and coccyx, ischial tuberosity, or heel), National Pressure Ulcer Advisory Panel-European Pressure Ulcer Advisory Panel-Pan Pacific Pressure Injury Alliance (NPUAP-EPUAP-PPPIA) classification system stage¹⁶; and treatment outcome. The Braden score was determined by the duty nurse. If a pressure injury did not develop, the lowest Braden score recorded was abstracted; if a pressure injury developed, the Braden score recorded before the pressure injury developed was used. Data were entered into a data entry template designed with EpiData (version 3.1, EpiData Association, Odense, Denmark) and retrospectively reviewed. 

If the information was not available on the electronic hospital record, the authors manually searched paper medical records according to the patient’s name and hospital number. If the necessary information could not be found, that case was excluded.

Statistical analysis. Because the statistical hypothesis was that the incidence of pressure injury in hip fracture patients varies with the month or season, hospitalization data were grouped into seasons defined according to the Gregorian calendar. First, pressure injury incidence was calculated with 95% CIs in each month and season (spring, summer, autumn, and winter). A periodogram was drawn to demonstrate the intensity of seasonal patterns. Odds ratios (ORs) and 95% CIs were calculated as estimates of risk. The month or season with the lowest pressure injury incidence was used as the reference category. 

A multivariate logistic regression model was used to calculate adjusted ORs for seasonal differences of pressure injury incidence. Again, the season with the lowest pressure injury incidence was used as the reference category. Diabetes mellitus, admission hemoglobin, admission albumin, length of surgery, and Braden Scale scores also were included in the logistic regression model as covariates. Statistical analyses were performed using Stata software, version 11.0 (Statacorp LLC, College Station, TX).

Patient characteristics. Of the 256 patients admitted with a hip fracture, 21 had missing pressure injury information, leaving 235 hip fracture patients (95 [40.4%] male, 140 [59.6%] female) for inclusion in the study. Mean patient age was 70.4 ± 10.5 (range 48–81) years. Thirty-three (33) patients (14.0%) had diabetes (HbA1c >6.5%), mean admission hemoglobin was 112.3 ± 13.9 g/L, and mean admission albumin was 32.6 ± 3.4 g/L. The length of surgery ranged from 45 minutes to 180 minutes. The mean Braden Scale score was 15.1 ± 2.5 (range 10–22) (see Table 1). 

Postoperative pressure injury. Thirty-one (31) patients developed 37 pressure injuries, an incidence of 13.2% (95% CI: 9.1%-18.2%). Pressure injuries developed 2 to 5 days after surgery; 18 (48.6%) were on the sacrum and coccyx, 12 (32.4%) on the ischial tuberosities, and 7 (18.9%) on the heel. Based on NPUAP- EPUAP-PPPIA classification, 30 pressure injuries (81.1%) were Stage 1, 5 (13.5%) were Stage 2, 2 (5.4%) were Stage 3, and none (0.0%) were Stage 4. All pressure injuries were managed with hydrocolloid dressings, and all patients were repositioned every 2 hours. All Stage 1 and Stage 2 pressure injuries were completely healed 4 to 10 days after occurrence; the 2 Stage 3 pressure injuries were partially (50% to 80%) healed at the time of discharge (13 to 21 days after surgery.) 

Braden Scale scores by season. The mean total Braden Scale score increased from June to August — the highest was during October (16.6 ± 3.7) and November (16.6 ± 3.5) and the lowest during June (14.2 ± 3.2); the moisture subscale was highest in October (3.5 ± 0.4). The remaining 5 Braden subscales showed no changes per month. Average Braden scores were highest in autumn (16.5 ± 3.4) and lowest in summer (14.4 ± 3.2 ) (see Table 1). 

Seasonal differences in the incidence of pressure injury. The incidence of pressure injury in hip fracture patients was lowest during the month of November (5.0%; 95% CI: 0.0%-24.9%) and highest during the month of June (22.7%; 95% CI: 7.8%-45.4%). A seasonal trend was observed as well, with incidence peaks occurring in summer (20.8%; 95% CI: 12.2%-32.0%; OR 3.3; 95% CI: 1.0-12.1; P = .025) when compared to autumn (see Table 1). 

Of all variables assessed, multivariate logistic regression analysis showed the Braden Scale score was the only independent risk factor for pressure injury incidence in patients with a hip fracture. Pressure injury incidence was highest in the summer season but was not statistically significant. After adjustment per the Braden Scale, the OR for the summer season was 1.537 (95% CI: 0.964-2.872) (see Table 2). 

A seasonal trend regarding pressure injury was noted in hip fracture patients, with incidence peaking in the summer (20.8%; 95% CI: 12.2%-32.0%). OR was 3.3 (95% CI: 1.0-12.1; P = .025) when compared with the study reference (autumn, 7.5%; the lowest incidence). The results of the current study support the validity of the Braden Scale score as a predictor of pressure injury risk for hip fracture patients; average total Braden Scale scores were lower in the summer, as were the Braden moisture subscale scores. These findings confirm what the authors had experientially observed and may be explained by the hot and humid summer environment. Although all wards in the authors’ hospital are equipped with air conditioners, the environment is still considerably warmer and more humid in the summer than during the other seasons. In June and July each year, the middle and lower reaches of the Yangtze River in China experience the rainy season with continuously overcast weather and very high air humidity. While the high temperature in the ward can be decreased by air conditioning, the high humidity levels are typically not fully addressed. 

Studies^17,18 have investigated the most suitable temperature and relative humidity in hospital buildings. One study in China¹⁹ found the measured temperature range acceptable to 80% of the people queried was 18.1˚ C to 25.5˚ C in the winter and 21.9˚ C to 28.4˚ C in summer. The American Society of Heating, Refrigerating and Air-Conditioning Engineers²⁰ suggests the relative humidity of these temperature comfort zones should range from 20% to 70%. In China, the relative humidity recommended for a hospital ward is 45% to 70% both in winter and summer.¹⁹ In the authors’ hospital, the relative humidity is always above 90% during the summer months because of the rainy season. The authors recommend placing a thermometer and hygrometer in the ward for hip fracture patients. If the temperature and humidity exceed a certain range, additional pressure injury prevention strategies should be provided. Specifically, measures to reduce skin moisture levels should be employed. For example, in China, hair dryers are used to keep patient skin dry: a randomized controlled trial by Fei-ying et al²¹ discussed the use of hair dryers for pelvic fracture patients and how this practice achieved a lower incidence of pressure injury. 

This study has limitations. First, the periodogram showed the highest pressure injury incidence occurred in the summer (P = .025), but multivariate logistic regression showed this was not statistically significant (P = .087). This result may be related to the small sample size. Second, temperature and humidity data in the ward were not collected retrospectively and as such cannot be investigated. Third, because this was a retrospective study, the levels of evidence may only ranked 3b.²² Future prospective studies with large samples can address these limitations.

Results of a retrospective review of patient data suggest the incidence of pressure injury in postoperative hip fracture patients may have seasonal variations. During the hot, humid summer environment in China, the incidence of pressure injury was higher than during colder, less humid, seasons. Braden Scale scores and the Braden Scale moisture subscale scores were lower during the summer than during the other seasons (P <.05 for Braden Scale score). Future prospective studies assessing  pressure injury prevention strategies during summer months, particularly with respect to decreasing ambient air temperature and humidity levels in wards in China, are warranted.

This work was supported by a grant from Key Medical Management Program of Nantong Municipal Science and Technology Bureau.

Dr. Chen is an Associate Professor, Nursing School of Nantong University, Nantong, Jiangsu, PR China. Dr. Zhu is a Deputy Chief Surgeon, Department of Hepatobiliary Surgery, The Second Affiliated Hospital of Nantong University and the First People’s Hospital of Nantong City, Nantong, Viangsu, PR China. Ms. Wei is a Deputy Chief Nurse, and Dr. Zhou is a Deputy Chief Surgeon, Department of Orthopaedics, Affiliated Hospital of Nantong University, Nantong, Jiangsu, PR China.

Please address correspondence to: Bin Zhu, MMed, Department of Hepatobiliary Surgery, The Second Affiliated Hospital of Nantong University and the First People’s Hospital of Nantong City, Jiangsu Province, PR China. 226001; pphss@126.com