Pressure injuries (PL) after surgery affect thousands of people worldwide, reducing physical health and quality of life [1]. The management of LP is expensive, a cost that could be reduced with appropriate prevention maneuvers [1,2].
Consequently, greater awareness in the prevention of LP has been observed in the last 10 years. The National Pressure Ulcer Advisory Panel (NPUAP) defines LP as “a localized injury to the skin and/or underlying tissue, usually over a bony prominence, as a result of pressure, or pressure combined with shear [3].
The development of a LP is a complex phenomenon, where both intrinsic and extrinsic factors are important [3]. The most important factor for increasing the risk of developing LP is reduced mobility [4]. LP can occur at any time when tissue is compressed, and the compression causes breakdown of the skin and underlying tissue.
The risk of LP increases when body weight is not evenly distributed over a support surface, or the tissue is poorly perfused [3].
Several areas of the body are most affected by LP. The most common sites after surgery are: the occipital area of the skull, the scapula, the elbows, but particularly the sacral area and the heels [5] (Fig. 1). The severity of LP varies, from erythema with intact skin, to tissue destruction, involving subcutaneous tissue, muscle and bone [6].
There are several different classification systems for evaluating LPs. The NPUAP and the European Pressure Ulcer Advisory Panel (EPUAP) have jointly published a classification system that divides LP into four stages: stage I, non-blanchable erythema; stage II, partial thickness skin loss; stage III, full-thickness skin loss; and stage IV, complete loss of tissue thickness [6] (Fig. 1).
To prevent LPs, patients at risk should be identified. The published literature has identified more than 100 potential risk factors for the development of LPs. Among these factors are mentioned: immobility, hypotension, sepsis, and diabetes mellitus [3,7].
Patients who have had surgery under general anesthesia are at risk of developing LP, as they are immobile and unable to change position [4]. LPs attributable to surgical procedures are therefore not uncommon [8]. Other perioperative factors are considered to increase the risk of developing these injuries, such as patient position and vasopressors [9,10].
Several tools have been developed to assess patients’ risk for developing LP. The most common are the Waterlow and Braden scales, both introduced in the 1980s. The Waterlow scale [11] assesses six factors that contribute to an increased risk of LP: weight/height, skin type, sex/age , continence, mobility, and poor nutrition.
Extra risk factors taken into consideration by the scale are: poor tissue nutrition, neurological deficit, and major surgery/trauma. A high Waterlow score indicates a higher risk of LP. The Braden scale [9] assesses the risk of skin breakdown in six domains: sensory perception, humidity, activity, mobility, nutrition and friction/shear.
A low Braden score indicates a high risk of LP. Currently, the Braden scale is the most widely used tool worldwide, with the exception of the United Kingdom, where the Waterlow scale is often preferred. None of the scales was designed for a specific group of patients [9].
Several studies have investigated the accuracy of these scales. A recent meta-analysis [10] concluded that the Braden scale has low predictive validity for assessing the risk of LP in surgical patients, and that it cannot be used alone in these patients. The validity and reliability of the Waterlow scale has also been investigated in a recent study [12].
The conclusion was that, due to the limitations of the scale, it should not be used alone, without clinical evaluation. A recent meta-analysis [13] investigated the predictive validity of these scales. A definitive conclusion could not be drawn due to the high degree of heterogeneity among the included studies. To date, no tool has been developed to assess the risk of developing LP after surgery.
This systematic review and meta-analysis aimed to summarize currently published data on perioperative risk factors for the development of postoperative LP in adult patients undergoing surgery under general anesthesia.
| Methods |
The review protocol was registered in the PROSPERO database before starting the study (CRD42019111877); https://www.crd.york.ac.uk/PROSPERO/) [14]. A bibliographic search was carried out based on the PRISMA guidelines.
> Eligibility criteria
Studies reporting on possible risk factors for postoperative LP in surgical patients operated under general anesthesia were evaluated for inclusion. Outcomes of interest were all perioperative variables reported in relation to subsequent postoperative LP. The perioperative phase was defined by the duration of admission for surgery, including on-call admission, anesthesia, surgery, and recovery.
> Sources of information and selection of studies
The following databases were searched for articles: MEDLINE, CINAHL, Embase, and the Cochrane Library. All searches were conducted by two authors on June 6, 2019, with the assistance of a research librarian.
The terms used for the search (in English) were: “pressure ulcer”, or “decubitus ulcer”, “ulcer”, and “surgery”, or “surgical procedures”, or “operative procedures”, and “factors”. risky". Medical Subject Heading (MeSH) terms , and subject headings and subheadings for each search term were explored.
A free text search was also conducted using the same terms. All grades of LP, all types of surgical fields, and all potential risk factors were eligible for inclusion. Only comparative studies were eligible.
The search was limited to publications within the last 15 years and published in English. Bibliographic lists of identified studies and previously published reviews were also explored.
The selection and evaluation of the articles was done using the Covidence program (https://www.covidence.org). Duplicates were removed and references were searched by title, abstract and full text. All stages were carried out individually by each observer. Disagreements were resolved by consensus. Where the full text of the study was not available, the corresponding author was contacted for access.
> Data collection
Data were extracted using a form designed by the authors. The extraction was performed individually by the two authors and then checked for accuracy. The variables sought were: title, country/year, type of study, type of surgery, objective, number of patients with LP, number of controls, factors associated with the development of LP, patient position and surface of the operating table. .
> Analysis of the results
A systematic review was conducted to identify variables for inclusion in the meta-analysis. Variables were summarized if they were reported in at least four articles. All variables reported by at least three articles, with comparable outcome measures, were analyzed.
The analyzes were carried out with the help of a biostatistician using the RevMan program, version 5.3 ( Cochrane Collaboration, Nordic Cochrane Centre , Copenhagen, Denmark). Meta-analysis was used when three or more studies reported on a specific variable.
The inverse variance method was used to calculate the results for continuous data, and the Mantel-Haenszel method for dichotomous data. The results of the meta-analysis are presented as weighted mean difference (MD) for continuous data, and hazard ratio (HR) for dichotomous data.
Heterogeneity was measured using I2 , t2, and c2 values. If heterogeneity was considered low, a fixed effects model was applied. The I2 value describes the percentage of total variation across studies, which is due more to heterogeneity than chance.
A value of 0% indicates no variation across studies, and 100% indicates maximum variation [16].
A t2 value was calculated for meta-analysis investigating differences in means using the random effects method. t2 investigates the between-study variance and is useful for comparison of heterogeneity, but the values depend on the size of the treatment effect [16].
The estimate of c2 is also an indicator of possible heterogeneity; a high value of c2 relative to its degrees of freedom indicates heterogeneity [16]. No specific threshold values were used, and the assessment of heterogeneity was an assessment of all calculated variables. For all hypothesis tests, a P < 0.050 was considered indicative of statistical significance.
> Quality evaluation
The quality of the studies was reviewed by two authors individually, using two different versions of the Newcastle-Ottawa Scale (NOS) [17], depending on the type of study evaluated. The versions used were those for cohort studies and for case-control studies. Disagreements were resolved by consensus.
The scale assesses three main categories: cohort selection (cases and controls), cohort comparability (cases and controls), and exposure/outcome. Several subcategories were also evaluated. Points were assigned for the evaluated categories. The total NOS score ranges from 0 (lowest grade) to 9 (highest grade). Studies with scores above the median (above 5) were classified as high quality [18].
| Results |
> Selection of studies
1454 citations were identified in the database. After removing duplicates and screening, 14 studies [19-32] met the inclusion criteria.
Forty-four studies were excluded based on evaluation of the full text; 20 had the wrong study design (mainly non-comparative studies), 16 had the wrong patient population, such as a mix of adults and children, and 8 studies were excluded for having wrong results, or for other reasons.
> Characteristics of the studies and quality evaluation
The 14 studies [19-32] included a total of 1903 new cases of LP. These studies included 7 prospective cohort studies, 3 retrospective cohort studies, and 4 retrospective case-control studies.
Only 1 study [25] reported LPs that developed in a specific anatomical location: the heels. Follow-up ranged from hours after surgery to 143 days.
Seven articles (19-21,22,24,28,30) did not specify the duration of follow-up. Only 65 studies [20,21,26,27,29,32] reported how patients were positioned during surgery. Neither study specified the type of mattress used during the operation.
In the quality assessment results, all studies scored above the median value and were classified as high quality. Six studies, [19-22,24,26] did not specify the absence of LP preoperatively. Two studies [24,27] did not report on the comparability of the patient groups. Four studies [22,26,29,30] did not report on completeness of data and any loss to follow-up.
> Meta-analysis results
The systematic review identified ten variables eligible for meta-analysis: age, sex, diabetes, serum albumin concentration, hemoglobin level, cardiovascular disease, respiratory disease, vasopressor use, duration of surgery, and steroid use.
All studies contributed to the meta-analysis, except that of Fred et al. [24], who did not report on the variables or type of data eligible for inclusion in any meta-analysis. To perform meta-analysis on the duration of the operation, the available data were recalculated using the method of Hozo et al. [33].
Of the 10 risk factors analyzed, 5 demonstrated statistical significance.
For the continuous variables, there was a significant association between the development of a LP and the longer duration of surgery, with a mean difference (MD) of 69.81 min (95% confidence interval [CI]: 2.36 to 137 .26) ( P = 0.04). Lower hemoglobin levels were also associated with an increased risk of LP: MD -7.94 g/L (-13.12 to -2.76) ( P = 0.003).
Relative to categorical data, several comorbidities were significantly associated with the development of LP. These were: diabetes (OR 1.49; 05% CI: 1.29 to 1.71; P < 0.001), cardiovascular disease (OR 2.24; 95% CI: 1.56 to 3.22; P < 0.001 ), and respiratory disease (OR 3.28; 95% CI: 1.89 to 5.71; P < 0.001).
The meta-analysis did not detect a significant association with the other factors investigated.
| Discussion |
This systematic review and meta-analysis sought to identify factors associated with the risk of LP in surgical patients. The following factors in the meta-analysis were significantly associated with LP: cardiovascular disease, respiratory disease, diabetes mellitus, low hemoglobin level, and longer duration of surgery.
Prevention methods for LP vary across hospitals and countries. Most hospitals use versions of the recommendations published by the NPUAP and EPUAP, which are general in design and do not take into account individual patient characteristics. The main focus is on the surgical table mattress and support aids such as pillows. Moving patients during surgery and regulating their temperature are also discussed.
There is little research on the degree to which positioning devices used in the operating room influence the development of LP, compared with individual risk factors among patients undergoing surgery. Little has changed over the past few decades; The mattresses used in the operating room are largely the same, the definition of LP is the same, and the scales used to measure the risk of LP are the same.
Recent studies [10,12], evaluating the reliability and validity of the Braden and Waterlow scales for surgical patients, have concluded that they are defective. A new tool is needed to evaluate the risk of LP in surgical patients.
The quality of previous studies on LPs has generally been low, and evidence on the effectiveness of supportive mattresses and pillows is scarce. Even though all articles published in the last 15 years were included, a relatively small number of them were eligible for inclusion in this review.
A better understanding of the risk factors for LP is essential because it may allow stratification of patients before surgery, and direct its prevention. In this study it was a surprise to find that sex was not associated with the development of LP, as had been reported in a previous meta-analysis [34]. A large number of studies were included in the present sex meta-analysis, and it is unlikely that an increased risk could have been missed.
It was also expected that serum albumin level could be associated with the development of LP, as an indicator of poor nutritional status, which may affect healing [35]. Only 3 studies reported on serum albumin; Conducting a meta-analysis with such a small number has its limitations, and this result should be interpreted with caution.
The main limitation of this review was the heterogeneity of the included articles. The population varied both in the type of surgery and emergency (acute/elective). The articles varied in their reports, given that not all studies included all stages of LP.
Numerous risk factors were reported, but relatively few meta-analyses could be performed. This was because few articles reported on the same variable, or because the variables were reported differently (categorical versus continuous data).
Other issues with the included articles were the possibility of LPs being present before surgery, loss of data, length of follow-up, and overall study design.
LP after surgery is generally preventable and at least partially predictable. Patients undergoing general anesthesia who have cardiovascular disease, respiratory disease, diabetes, a low hemoglobin level, or who are expected to have a prolonged duration of surgery are at risk.
It should be possible to develop more sophisticated risk scoring tools, to allow high-risk patients to receive better prevention measures. This research should also stimulate a greater focus on preventive measures to offer optimal protection to high-risk patients during surgery.
