Factors Associated With Pressure Injury Development in Older Hospitalized Patients: A Prospective Descriptive Study

A pressure injury (PI) results from local injury to the skin and/or underlying tissues due to unrelieved pressure or pressure with shear or friction.^1-3 PIs are usually not the primary cause of hospitalization,⁴ but PIs adversely affect patient well-being and nurse workload. Hospital-acquired pressure injuries can create a substantial financial burden on a health care system.⁵

Pressure injuries frequently occur in patients who are immobile, with more than 72% of PIs occurring in older adults.⁶ Advancing age can result in skin changes and, when coupled with reduced mobility and multiple comorbidities, increases the risk of PI development.⁴ Multiple chronic health conditions must be taken into consideration, including nutrition,⁷ older age, comorbidities,⁸ diabetes, peripheral arterial disease, immunodeficiency, poor nutrition associated with frailty⁹ and older populations,^9-11 corticosteroid therapy, and smoking status.¹² In a randomized controlled trial, Latimer et al¹⁰ reported that for every additional comorbidity, older adults with limited mobility are 22% more likely to develop a PI within 36 hours of hospitalization. Therefore, there is a need to identify risk factors for PIs in older patients so that PI development can be avoided. Prior PIs can increase the risk of additional PIs,¹³ but the exact significance of prior PIs is yet to be elucidated.¹⁴

Many studies have focused on the prevalence and incidence of PIs, but many were cross-sectional or retrospective.^15-17 The aim of the current prospective descriptive study was to identify factors associated with PI development in older hospitalized patients.

Study design and setting. This was a prospective descriptive single-center study conducted at the Jiangsu Provincial Hospital, which is a research medical hospital with more than 4500 beds in east China. Twelve wards accommodate older patients and are collectively referred to as the Geriatric Department. The study was approved by the ethics committee of the hospital (no. 2018-SR-330). The study was performed in August and September 2018. Informed consent was waived by the ethics committee because it was believed that some in vitro measurements and blood tests necessary for standard diagnosis and treatment were deemed harmless and that informed consent was not required. Nevertheless, all were verbally disclosed at the time of enrollment.

Participants. Bedridden patients who were 65 years and older were included in the study. Patients for whom a physical examination could not be completed and those with skin or vascular disease were excluded.

Standard care. Patients who were bedridden and at risk of PIs were all treated similarly.

Standard care was provided according to National Pressure Injury Advisory Panel guidelines.¹³ Every patient was provided with a pressure-relief pillow, which is a small, triangular relief pad that supports the patient’s back; it is also called a turning pillow or medical care turning over pad. Caregivers were informed of the preventive measures for avoiding rubbing and friction. Pressure injury risk was assessed for each patient by registered nurses using the Braden scale¹⁸ and the Ohura-Hotta scale.¹⁹ If a patient had a Braden scale score ≥ 17, PI risk was assessed on a weekly basis. If a patient had a Braden scale score between 13 and 16, an air-filled mattress was provided within 4 hours and PI risk was assessed every 3 days. If the patient had a Braden scale score ≤ 12, PI risk was assessed every day; these patients were also placed on a pressure-redistribution mattress.

During each shift, the staff nurse inspected the skin at all contact points and moved and repositioned patients every 2 to 3 hours throughout the day and night. Each patient was cared for while alternating between the supine and lateral positions. Risk assessment was repeated according to the study protocol above or if marked changes occurred in the patient’s clinical status. When a PI developed, the stage was determined,³ wound care was addressed, dressings were changed as necessary, and wound changes were monitored; all measures were performed by the appointed ward nurse according to clinical practice guidelines.³ With the exception of nasogastric feedings, patients were required to remain in a semirecumbent position with the head of the bed raised approximately 30 degrees during and 30 minutes after feeding. Because nearly all patients were in the supine position and bedridden, the authors did not collect data about the head-of-the-bed angle.

Pressure injury risk factors. Pressure injury risk factors collected and analyzed for this study were identified by a panel of experts. The panel included a nurse expert in wound care, a nurse expert in research methodology, a physician, a geriatrician, and a statistician. The variables were divided into two areas: intrinsic properties of the patient and extrinsic factors. The intrinsic variables included PI risks (Braden and Ohura-Hotta scale scores); body mass index (BMI) (underweight, < 18.5 kg/m²; healthy weight, 18.5–24.99 kg/m²; overweight, 25–29.99 kg/m²; and obese, ≥ 30 kg/m²)¹⁷; hemoglobin (Hb), erythrocyte (RBC), albumin (ALB), blood glucose (BG), and leukocyte (WBC) levels throughout hospitalization (ALB, Hb, and other laboratory indicators were collected from the patients’ blood sample on the day of admission or the next morning, and were then reviewed weekly); and skinfold thickness. The extrinsic variables included PI preventive measures (air-filled mattress and dressings) and pressure to the sacrum.

Evaluations. An electric shifter (Maxi Move; ArjoHuntleigh, Sweden) was used to measure patient weight, and a measuring tape was used to measure height. BMI was calculated as the weight in kilograms divided by the height in meters squared. To measure pressure to the sacrum, the patient was assisted in the supine position. The piezometer of the portable contact pressure meter was placed on the sacrum, and the display screen was directly read to record the pressure value. The patient was then assisted in the lateral position. Skinfold thickness was measured by gently lifting the subcutaneous fat with the left thumb and index finger, and the portable contact force tester was held in the right hand; the 2 ends of the tester were inserted into the pinched skin on both sides to the bottom, and the measurement was obtained. The corners were confirmed to be in close contact with the skin.

Scales and instruments. The Braden scale includes 6 subscales: sensory perception, moisture, activity, mobility, nutrition, and friction/shear. The friction/shear subscale is scored from 1 to 3, and the remaining subscales are scored from 1 to 4. Lower scores indicate greater risk. The lowest total score is 6, indicating the highest risk of PI development. Two prospective studies of predictive validity were completed to determine the Braden scale’s sensitivity and specificity. Reliability ranged from r = .83 to r = .94 for nurses’ aides and licensed practical nurses; when used by registered nurses, the reliability increased to r = .99. Predictive validity was calculated for each cutoff point of the scale. Using a cut-off point of 16, sensitivity was 100% in both studies. Specificity ranged from 64% to 90%.²⁰

The Ohura-Hotta scale was created by the Japanese Health Department based on the Braden scale and was simplified into 4 assessment items (ability to autonomously change body position, pathologic bone protrusion, edema, and joint atrophy).¹⁹ Edema and joint atrophy items divided into 2 rating levels, the ability to change position autonomously and pathologic bone protrusion were divided into 3 grades. The value range of the 4 evaluation items was 0 to 3, with a full score of 10 (7 to 10 being high risk, 4 to 6 being medium risk, and 1 to 3 being low risk). The Cronbach's α reliability coefficient and Spearman’s correlation coefficient of the Ohura-Hotta scale were 0.694 and 0.444 to 0.697, respectively, showing a significant correlation and good internal consistency reliability. The cumulative contribution rate of variance of the two common factors generated by factor analysis was 65.577%, and each item had a high factor load.²¹ The areas under the curve for the Braden scale and Ohura-Hotta scale were 0.814 and 0.794, respectively.

The Barthel score was used to measures independence in performing ADLs.^6,22 The Barthel score includes 10 items (feeding, bathing, grooming, dressing, bowel incontinence, bladder incontinence, toilet use, transfers, mobility, and stairs), with the total value ranging from 0 to 100. A higher score indicates greater independence in performing ADLs.

Outcome. The primary outcome of this study was the occurrence of PIs during hospitalization, which was defined as the total number of PIs that occurred 1 month after enrollment, including PIs that were present at enrollment. A PI was defined as damage to the skin and underlying tissues from pressure, friction, shearing, or a combination of these factors.²³

Data collection and follow-up. Data collection and examinations were performed once a week for 1 month of hospitalization. Participant information was obtained by a registered nurse, who completed a series of PI training courses. Blood test results were obtained from standard blood tests, and body measurements were obtained by nurses using with measuring tools. Comorbidities, age, sex, weight, height, BMI, sacrococcygeal pressure skinfold thickness, Ohura-Hotta scale score, Braden scale score, and Hb, RBC, BG, WBC and ALB levels were recorded and calculated by the researchers.²⁴ The nurses checked for PIs every day at the shift change, including day to evening shift (at 6 PM), evening to night shift (at 1AM), and night to day shift (at 8AM). Nurses who identified PIs were responsible for recording adverse events and monitoring the PIs. New PIs were immediately reported to the appointed ward nurse and documented on special data collection sheets. When a PI developed, the stage was determined,³ wound care was addressed, dressings were changed as necessary, and wound changes were monitored; all measures were performed by the appointed ward nurse according to clinical practice guidelines.³

All study variables recorded in the questionnaires over the 1-month period were included on one datasheet. To ensure the reliability of PI grading and care, the dedicated nurses had taken part in drafting the protocol. They had tested the materials and methods in a pilot study of 86 patients conducted in May 2018 (unpublished data).

Statistical analysis. The data were analyzed using SPSS 19.0 (IBM Corp). The continuous variables were tested for normal distribution using the Kolmogorov-Smirnov test. The continuous data with a normal distribution are presented as means ± standard deviations and were analyzed using Student t-test; those with a non-normal distribution were presented as medians (ranges) and were analyzed using the Mann-Whitney U-test. The categorical data were presented using n (%) and were analyzed using the chi-square test. The factors associated with PIs were identified using univariable and multivariable logistic regression analyses. Two-sided P values < .05 were considered statistically significant.

Of all 369 patients who were admitted to the Geriatric Department, 144 (39.0%) met the inclusion criteria for the study. A total of 124 patients (33.6%) were enrolled; 4 patients (1.1%) did not complete the study (Figure). Therefore, 120 (32.5%) patients were included in the analyses. Table 1, Part 1 and Table 1, Part 2 present the characteristics of the participants.

Occurrence of PIs. The PIs were categorized into 6 stages: suspected deep tissue injury (DTI); stages 1, 2, 3, and 4; and unstageable.²⁵ PIs occurred in 20 participants (16.7%), and there were 21 PIs total (1 patient had 2 PIs). Six of the 20 participants had PIs originally, and 12 of the 14 new PIs occurred during week 4 of hospitalization. The PIs were located in the sacral area in 10.0% of patients (n = 12), the heel in 2.5% (n = 3), and in other areas in 4.2% (n = 5; 2 on the dorsum, 1 on the buttocks, 1 on the ankle, and 1 on the auricle and bridge of the nose). Among the PIs, there were 11 that were stage 1, 8 that were stage 2, and 2 that were stage 3. There were no PIs in stage 4, unstageable, or deep tissue injury.

Compared with the patients without PIs, those with PIs had lower Hb levels, thinner skinfold thickness, higher sacrococcygeal pressure, higher Ohura-Hotta scale score, lower Braden scale score, lower BMI, and lower ALB (all P < .05) (Table 1). There were no statistically significant differences related to age and sex. Patients with complicated pulmonary diseases had more PIs, which may be related to pulmonary infection.

Factors associated with PIs. The univariate analyses showed that several variables were associated with the occurrence of PIs (Table 2). The multivariate analysis showed that higher sacrococcygeal pressure (odds ratio [OR] = 1.06; 95% confidence interval [CI], 1.00-1.11; P = .045), lower Braden scale score (OR = 0.66; 95% CI, 0.48-.92; P = .012), and lower albumin levels (OR = 0.66; 95% CI, 0.49-0.89; P = .006) were associated with the occurrence of PIs (Table 2).

In this study, most PIs occurred during the fourth week of hospitalization. Although the prevalence of stage 2 and higher PIs was nearly 50.0% of patients, 30.0% (n = 6) had already developed PIs before the beginning of this study, and 21.4% of the newly developed PIs (n = 3) healed within 1 week. Thus, PIs can be mild in severity, even in older patients, with timely detection and discovery as well as appropriate treatment.

Every patient in the current study had caregiver who was informed of the preventive measures for avoiding rubbing and friction. These caregivers assisted the nurses to turn the patient regularly and maintain posture after turning. After PI development, the caregiver would report to the nurse, and the patient would be turned more often under the guidance of the nurse.

Of all the intrinsic factors in this study, only the Braden scale was a significant risk predictor. This is consistent with most studies that have evaluated the Braden scale, which is considered the gold standard.²⁶ Screening tools with high sensitivity, especially for older adults, would be optimal compared with screening tools with high specificity, which includes the Norton scale and the Braden scale.²⁶ The currently used tools that assess the risk of PIs rarely evaluate the accuracy of predictability in older adults, and nurses have raised the issue that the screening tools for elderly patients are not appropriate to predict the occurrence of PIs.^27,28

The Ohura-Hotta scale has been shown to have greater reliability and validity as a PIs assessment indicator in older Japanese and Chinese patients.^29,30

A multicenter, cross-sectional study was conducted with 34 home-based geriatric support service providers located in 5 Japanese districts. The study included 69 participants (30 had a PI and 39 did not).³¹ The Braden and Ohura-Hotta scales served as comparison tools. The areas under the curve for the Braden scale and Ohura-Hotta scale were 0.814 and 0.794, respectively. Because of the small sample size in this study, the Ohura-Hotta scale could not be confirmed to assess and predict the risk effectively in older adult patients or to identify the most at-risk patients for intervention.

Another intrinsic variable that was associated with the risk of PI development in this study was skinfold thickness, which has been shown to influence PI development. The thickness of the sacrococcygeal fat layer represents a person’s overall nutritional status, with greater thickness indicating good local microcirculation, which can alleviate pressure damage to the skin and subcutaneous tissues to varying degrees.³²

This study was undertaken in one institution and had a small sample size (N = 120). Patients were followed-up for only 1 month. The study design did not allow the determination of the prevalence of PIs. Furthermore, the length of stay of the enrolled patients was not collected and analyzed in its entirety, and no relevant data were collected about head-of-the-bed angle. Finally, the role of caregiver in the hospital may not be the same as in other health care institutions in other countries. Future studies are warranted to address these limitations.

The aim of this prospective descriptive study was to describe the incidence, stage, and predictors of PIs in a group of older hospitalized patients in China. The sample consisted of 124 bedridden patients aged 65 years and older. A 16.7% incidence of PIs was found within 1 month of hospitalization. Results suggest that higher sacrococcygeal pressure, lower Braden scale score, and lower albumin level were independently associated with the occurrence of PIs in these patients. These indicators may help identify older patients at risk for PIs. Additional studies are needed.

Ms Li and Ms Mei are nurses, Department of Geriatrics, the First Affiliated Hospital with Nanjing Medical University, Nanjing, Jiangsu, China. Ms Lin is a professor, Department of Nursing, the First Affiliated Hospital with Nanjing Medical University, and a professor, School of Nursing, Nanjing Medical University, Nanjing, Jiangsu, China. Ms Zhao is a nurse, Outpatient Treatment Room, the First Affiliated Hospital with Nanjing Medical University, Nanjing, Jiangsu, China.

Address all correspondence to: Zheng Lin, MS, Department of Nursing, the First Affiliated Hospital with Nanjing Medical University, Nursing Research Center, Nanjing Medical University, No. 300 Guangzhou Road, Nanjing 210029, Jiangsu, China; tel: +86-18915971023; email: linzheng100@163.com.

This study was funded by the Jiangsu Province Cadre Health Care Project Funding Agency
(grant number BJ17016).