A Retrospective Comparison of Patient and Pressure Injury Data at Two Time Points in a Regional Hospital in France

Although the prevention and treatment of pressure injuries (PIs) historically may have been viewed as a nursing problem, their incidence is considered a useful quality-of-care as well as patient safety indicator¹ and deserves ongoing attention and research.

The reported overall prevalence of PI varies among countries. In the United States, prevalence ranges from 2.2% to 23.9%.²; in Canada, PI prevalence is estimated to be between 15.1% and 25.1%.³  In a recent meta-analysis of 39 studies from all over the world (total patient sample, 2 579 049) conducted by Li et al,⁴ the pooled prevalence was 12.8% (95% confidence interval, 11.8-13.9), and the most frequently occurring stages were stage 1 (43.5%) and stage 2 (28.0%).

Comparing the prevalence of patients with PI over time at different time points suggests that it is difficult to improve worldwide prevalence rates. Olivo et al⁵ conducted a cross-sectional study, using 2 separate designs at 2 separate timepoints (2010 and 2015), based on the classification recommended by the European Pressure Ulcer Advisory Panel (EPUAP). They found that the prevalence of PI in hospitals was 19.5% in 2010 and 17% in 2015. The percentage of patients with PI present on admission were 9.60% in 2010 and 9.42% in 2015, and hospital-acquired PI rates were 5.08% in 2010 and 5.87% in 2015. Older age, comorbidities, and a total Braden score ≤ 16 were positively associated with the presence of a PI (P < .05).⁵ Tariq et al⁶ found pressure ulcer prevalence to be 10.4% in 2018 versus 6.4% in 2013. The 2018 hospital-acquired pressure ulcer prevalence was 1.8% versus 2% in 2013, and the weighted average cost (per case) was approximately US $8035.⁶ Kirkland-Khyn et al⁷ performed a descriptive study involving prospective and retrospective data collected from the National Data for Nursing Quality Indicators. The average community-acquired PI prevalence and hospital-acquired PI prevalence were, respectively, 6.6% and 0.8% for 2015, 6.0% and 1.5% for 2016, and 6.9% and 0.9% for 2017. The average length of hospital stay for 2013 and 2017 was 38.9 days and 10.5 days, respectively.⁷

Pressure injury severity is evaluated in different ways. The key components of staging include thickness of the skin loss and other characteristics of the injury; more specifically, its location, size, presence of drainage (amount, color, and odor), presence of viable tissue in the wound, and presence of  undermining/tunneling. The National Pressure Injury Advisory Panel (NPIAP) developed a widely accepted classification system⁸; the authors’ hospital uses this system to evaluate PI characteristics.

According to an extensive review of the literature by Sullivan and Schoelles,⁹ a multifactorial strategy is the most successful for PI prevention. This approach involves simplification and standardization of interventions and documentation as well as a multidisciplinary team and clinical leadership (at the administrative level, audits and evaluations, continuous training, and the presence of expert centers).⁹

Steps for PI prevention have been defined by the NPIAP in a multimodal therapeutic tool.¹⁰ This tool includes the implementation of risk assessments; inspection of the skin; nutritional, diagnostic, and therapeutic interventions; patient repositioning and mobilization; and patient and family education strategies.¹⁰

The large number of studies on this subject demonstrates the persisting severity of the problem. The purpose of this study was to assess patient and PI information/characteristics (eg, age and PI severity profiles) to determine whether there were any significant differences between the 2 chosen years related to the facility’s evolving multidisciplinary approach (eg, changes in support surface and nutrition protocols) as well as factors that could impact PI prevalence.

With regard to the wide variety of terms used in the literature (eg, pressure ulcers, pressure injuries, bedsores, decubitus ulcers), the authors will use “pressure injuries (PIs)” when describing related lesions.

Setting and patients. Data were extracted retrospectively from the charts of patients who were admitted to 1 of 22 units of an 846-bed state hospital serving a population of 120 000 persons on October 20, 2009, and on October 24, 2013. Patients of all ages on all wards, except pediatrics and obstetrics, were included in the study. The hospital has a wound care team consisting of a practician with relevant training and extensive experience with PIs as well as a number of specially trained wound care nurses, nutritionists, physiotherapists, and physical exercise specialists dedicated to offer expertise to all departments and local long-term care facilities and nursing homes as needed. Wound care personnel were almost unchanged in the 2 time periods, and the facility follows a multidisciplinary wound care approach. Skin checks were performed on admission and subsequently during daily nursing rounds; the Braden scale chart was completed during the skin checks.

Procedure. As part of the hospital protocol, the unit manager and a unit nurse from each unit conduct an audit using a questionnaire for all patients present at 8 AM. During the audit, a set of data are extracted from the patients’ electronic charts and collected into a collective Excel file. For this study, the data from the 2 study dates were retrospectively extracted from the Excel files in November 2017. The variables included patient and PI characteristics (including age, sex, and the presence of PI [including the number and location for each patient]), hospital-acquired/community-acquired PIs, PI stage, Braden Scale score on admission, mattress type used for every patient, frequency of daily pain evaluation, type of feeding, use of oral nutritional supplements, total number of patients receiving physical and occupational therapy sessions, and patient mobility status. Data from patients with unstageable or deep tissue injury PIs as well as patients with mucosal pressure injury were not extracted. All patient identifiers were removed to assure anonymity. Data from patients with multiple PIs of different levels of severity were included.

Data analysis. Erroneous, incomplete, or missing data were eliminated; remaining data were transferred to an SPSS database (IBM) for further analysis. Descriptive statistics were used and the following variables were compared between the 2 time periods: PI prevalence (percentage of patients with a PI), total number of identified PIs per time period, sex distribution, mean age, distribution of PI stage (stages 1, 2, 3, and 4), PI location distribution (heels, sacrum, ischions, and other), and Braden Scale score (< 12 or ≥ 13). Distribution of the types of mattresses and cushions avalable during the 2 different time periods (standard mattress, waffle mattress, Tempur mattress [Tempur-Pedic, Lexington, KY], air mattress, gel cushion, or other) as well as  the distribution of patients receiving enteral or parenteral feeding, feeding via a percutaneous gastrostomy or a nasogastric tube, and use and daily number of oral nutritional supplements were also compared. Finally, the percentage of patients with the capacity for chair-to-bed transfer,  to move indepedently in bed, or being unable to reposition were evaluated. Qualitative data were compared using a continuous parameter between the 2 dates using a t-test for independent variables. The remaining variables studied were compared using Pearson’s chi-square method. P <.05 was considered significant.

Because this study involved 2 specific days at the hospital, no sample size was determined and no power analysis was performed.

Ethical considerations. Institutional Review Board of the Vichy District Hospital, France, approval for the study was obtained, and patients provided oral informed consent at the time of data collection.

The 2009 data included 515 patients (mean age, 70 years; range, 10–102). Of those, 281 were women (mean age, 77 years; range, 22–102). Ninety-seven (97) patients had at least 1 PI (19%), 75 patients (73%) had at least stage 1 or 2, and 22 patients (27%) had at least 1 stage 3 or 4 PI, for a total of 160 PIs. In 48 patients (49%), PIs were hospital-acquired; in 49 (51%), they were community acquired and present on admission. In 2013, 532 patients (mean age, 72 years; range, 23–103) were admitted on the day studied. Of those, 288 (54%) were women (mean age, 79 years; range, 23–103) and 84 patients (15.8%) had 133 PIs. Of these, 49 (58%) were hospital-acquired and 35 (42%) were present on admission (Table 1).

No statistical difference was noted in PI rates, location, and occurrence before or after admission between the 2 dates (Table 1), but a significant difference was observed in the mean age of hospitalized patients (Pearson χ²= 36.44; P = .014). Age distribution differed for patients ≤ 59 years (2009: 86 of 515 patients [16.7%] vs 2013: 63 of 532 patients [11.84%]), but remained relatively equivalent for patients between 60 and 79 years (2009: 209 of 515 patients [40.58%] vs 2013: 206 of 532 patients [38.72%]) and differed for patients ≥ 80 years (2009: 220 of 515 patients [42.72%] vs 2013: 263 of 532 patients [49.44 %]). The severity of injuries was significantly different between the 2 dates (2009 and 2013, respectively): stage 1 data included 48 patients (49%) versus 29 patients (35%); stage 2: 27 (28%) versus 25 (30%); stage 3: 18 (19%) versus 16 (19%); and stage 4: 4 (4%) versus 14 (16%) (Pearson’s χ² = 42.16; P = .010). No statistical difference was noted between the 2 dates with regard to Braden Scale scores.

Significant differences were noted regarding the mattress type used. In 2009, 106 patients (20.6%) were on a waffle mattresses versus 2 (0.4%) in 2013 (Pearson’s χ² = 157.74; P <.001). Tempur mattresses with memory foam were used for 316 patients (61.4%) in 2009 versus 427 (78.4%) in 2013 (Pearson’s χ² = 157.74; P < .001), and in 2009, 18 patients (3.5%) were on an air mattresses compared with 50 (9.4%) in 2013 (Pearson’s χ² = 157.74; P <. 001). Very few patients were placed on standard and water mattresses in both time periods (Table 2).

Significant differences in the frequency of pain evaluation during the morning audit (normally it should be done once daily, every morning) using the Visual Analogue scale in French (Echelle Visuelle Analogique [EVA] scale)¹¹ were noted between the 2 dates. In 2009, 58 patients were assessed versus 157 patients in 2013 (Pearson’s χ² = 8.88; P = .003).

The use of oral nutritional supplements differed significantly, with 100 of 515 patients (19.41%) versus 170 of 532 patients (31.95%) receiving nutritional supplements in 2009 and 2013, respectively (Pearson’s χ² = 21.49; P < .001) (Table 3). Additionally, significant differences were found in nutritional strategy for patients between the 2 dates. In 2013, parenteral nutrition was used more frequently than percutaneous endoscopic gastrostomy or a nasogastric tube, which were used regularly in 2009 (Pearson’s χ²= 16.70; P = .001) (Table 3).

No significant difference was noted between the 2 time points concerning the number of patients receiving physiotherapy sessions (169 in 2009 and 179 in 2013; Pearson’s χ² = 0.083) and occupational therapy sessions (29 in  2009 and 42 in 2013; Pearson’s χ² = 1498). The number of physical activity sessions differed between the 2 time points (28 sessions in  2009 and 54 in 2013; Pearson’s χ² = 6.71; P = .01).

Significant differences were found between the 2 dates with regard to the number of patients who could move independently from chair to bed (227 in 2009 and 308 in 2013; Pearson’s χ² = 13.46, P < .001) and who could reposition independently in bed (335 in 2009 and 426 in 2013; Pearson’s χ² = 16.79, P < .001) (Table 4). No significant difference was noted between the 2 dates in the number of patients who were unable to reposition independently (Table 4).

In this retrospective comparison study, the authors assessed PI prevalence and patient and PI information/characteristics to determine whether there were any significant differences between the 2 chosen years related to the facility’s evolving multidisciplinary approach (eg, changes in support surface and nutrition protocols) as well as factors that could impact PI prevalence.

Patient average age was significantly higher in 2013 than in 2009, whereas the PI prevalence rate decreased from 2009 to 2013 but not significantly (19% to 16%). However, this prevalence rate is comparatively better than in other countries. The total prevalence rate of PI in the United Kingdom was estimated in 2014 as being 8.4%.¹² In a review article from 2013 examining the rates of PI in Scandinavian countries, the average  prevalence in Norway was 17% (range, 4.8–29%), in Ireland 16% (range, 4–37%), in Denmark 15% (range, 2.2–35.5%), in Sweden 25% (range, 0.04–42.7%), and in Iceland 8.9%.¹³ In a study from Ontario published in 2017 with data accumulated from 2010 to 2013, the overall prevalence of PI was approximately 13% and highest in the complex continuing care setting.¹⁴ In a systematic review from 2019¹⁵ on the prevalence of PI in European countries with 79 articles included, the median prevalence was 10.8% (standard deviation, 7%; range, 4.6–27.2%); the highest PI prevalence reported was from the Netherlands (27.2%; n = 17 494 participants), and the lowest was reported from Finland (4.6%; n = 629 participants). Almost 32.4% (n = 151,195) of the pressure injuries were category 1, and the most common site for PIs was the sacrum.

In another meta-analysis with 39 studies in an exhastive literature research from 2008 to 2018 including studies with observational, cross-sectional, or longitudinal designs, the pooled prevalence of 1 366 848 patients was 12.8% (95% CI, 11.8-13.9%); the pooled incidence rate of 681 885 patients was 5.4 per 10 000 patient-days (95% CI, 3.4-7.8), and pooled hospital-acquired PI rate of 1 893 593 patients was 8.4% (95% CI, 7.6-9.3%).⁴ The most frequent stages were stage 1 (43.5%) and stage 2 (28.0%). The most affected body sites were sacrum, heels, and hip. Significant heterogeneity was noted across some geographic regions.

In a prospective study conducted by the Social Security Service of the Paris region among 21 998 hospitalized patients, the overall prevalence of  PI was 9.1%; 47% occurred during hospitalization, with stages 3 and 4 representing 32% of the total prevalence rate.¹⁶

In this study, almost half of the PIs were hospital-acquired despite the quality improvements in the multidisciplinary treatment strategies.

The use of specialized devices (ie, mattresses) increased sharply between 2009 and 2013. It is the authors’ opinion that foam mattresses rather than standard mattresses in the hospital should be used among people at a high risk of PI development. In a recent international consensus on PI preventive interventions by risk level for critically ill patients, it was proposed that moderate- and high-risk patients should receive the following: a reactive mattress support surface and a heel offloading device; high-risk patients should also receive preventive dressings (sacral, heel, trochanteric), an active mattress support surface, and a pressure-redistributing cushion for sitting.¹⁷

The findings of the current study regarding the location of PI are similar to the literature. In a multinational prevalence study,¹⁸ the prevalence of sacral PIs was 28% and for heel PIs was 23.6%; more than 70% of the injuries observed were stage 1 and 2. In this study, the majority of PIs in both time periods (46% and 44%, respectively) were on the heels and stage 1 or stage 2 (76% and 65%, respectively).

The implementation of consultations with a wound care specialist may reduce PI and improve quality of life. For example, after multidisciplinary management of inpatients was instituted in the Ontario region of Canada, the PI rate decreased from 5.8% to 1.6% over 4 years.¹⁹

The establishment of multidisciplinary measures and comprehensive care can be considered to be an effective measure to lower PI rates in health systems. Their implementation produced an 89% reduction in full-thickness hospital-acquired PIs in a US tertiary center with 400 beds.²⁰ In another study using the implementation of a multidisciplinary approach with a continuous update of new skin care products in the context of hospital-acquired PI prophylaxis, these interventions proved to be effective in decreasing PI prevalence from 11.7% to 2.1%.²¹

Regarding the current study, the authors believe that changes in multidisciplinary interventions (nursing, nutrition, physical therapy, and physical activity) as well as the continuous improvements in materials for PI prophylaxis has kept the rate stable despite the increase in the mean age of hospitalized patients.

Dietary management was much more rigorous in the authors’ facility in 2013 after the publication of numerous studies about the importance of nutritional status as a risk factor of PI and the use of supplements as an integral part of their prevention and management.^22-28

According to a recently updated Cochrane review,²⁹ evidence regarding medical nutritional therapy for the prevention of PI is inconclusive. In terms of nutritional policies at the authors’ institution in 2013, more patients received complementary nutritional support than in 2009, at which time more patients had a gastrostomy and/or nasogastric tube. The high rate of complications associated with gastrostomy has led to a reduction in its use.³⁰

No statistically significant difference was found between the 2 dates concerning the number of physiotherapy and occupational therapy sessions, but more physical activity sessions were reported in 2013. These practices may not significantly reduce the rate of hospital-acquired PI, but it is believed that such therapies can improve patient quality of life and clinical indices (functional fitness, strength, endurance, balance, flexibility, ability to perform activities of daily living, falls prevention, and alleviation of depression and incontinence symptoms).^29,31

Finally, the current study found that daily pain evaluation using the EVA scale occurred more frequently in 2013, but the overall rate was low and it should include all patients. It seems that there is a serious need for continuous education to sensitize paramedical personnel to the need to constantly evaluate pain in the patient population. More logistical and organizational resources are needed to implement these important changes in PI assessment and management.

Clinicians are constantly confronted with complex problems and issues that do not have simple solutions. Although PI prevention is far from simple, concrete actions can be taken. This study as well as many of those discussed in this article provide concrete and helpful practices, including documentation to improve surveillance capabilities and system-level team efforts to improve the quality of care and patient safety.³²

There are many competing priorities each day, but with the current state of the health system, any action aimed at conserving health care resources must be considered.³³

There are several limitations to the current study. First, 2 time points may or may not indicate a trend; numerous extraneous variables may have affected the results. Second, no information regarding the outcome of these PIs and the patients, including their physical and nutritional status, comorbidities, or mortality rates, were available. Third, the 2-point character of the study limits extrapolation of the results. Fourth, the results of this single-center study are limited to the nature of the specific center (ie, its semi-urban, peripheral location; age distribution of the local population; and the demographics of the study’s geographic location).

This retrospective, single-center, 2-time-point study showed a significant difference in the age of the hospitalized population between 2009 and 2013 as well as adaptations of care associated with PI prevention and management, such as the use of nutritional supplements. The rate of PIs and their anatomical locations were not significantly different, but there were significantly more stage 4 PIs in 2013 compared with 2009.

More prospective, multicenter studies are needed to evaluate existing strategies for the prevention and treatment of PI in a national and international setting. Such studies may help demonstrate the effects of the continuous improvements in therapeutic protocols of care  to offset the therapeutic challenges of an aging and frailer patient population.

Dr. Kosmadakis is a nephrologist, Ms. Martinez is a wound care nurse, Ms. Andre is a wound care nurse, and Dr. Da Costa Correia is a wound specialist, Jacques Lacarin Vichy General Hospital, Vichy, France. Address all correspondence to: George Kosmadakis, MD, Pole Metabolique-Centre Hospitalier Vichy, 54 Boulevard Deniere 03207 Vichy, France BP 2757; tel: +33(0)470973333; fax: +33(0)47972225; email: George.kosmadakis@gmail.com.