Evidence on the effect of position change on pressure ulcer among hospitalized adult patients in Ethiopia: systematic review and meta-analysis

Background Pressure ulcers (PU) aﬀect millions of people worldwide and always occur over bony areas of the body where pressure and tissue distortion is greatest. The national pooled prevalence of pressure ulcer remains unknown. Hence, this meta-analysis aimed to determine the effect of position change on pressure ulcer among hospitalized clients in Ethiopia. Methods Studies were retrieved through search engines in PubMed, Scopus, WHO afro library, Google Scholar, Africa journal online, PsycINFO and web-science following the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P). Analysis was done using STATA version 14 software. Heterogeneity between-study was checked using the I 2 and examined a potential publication bias by visual inspection of the funnel plot and Egger’s regression test statistic. The random-effect model was fitted to estimate the summary effects, odds ratios (ORs), and 95% confidence interval (CIs) across studies. Results Out of the reviewing 401 studies, 7 studies fulfilled the inclusion criteria and were included in the meta-analysis. The estimated pooled prevalence of pressure ulcer in Ethiopia was 11.7% (95% CI (7.28, 16.13%)). Based on the subgroup analysis, the estimated magnitude of pressure ulcer was 15.89% (95% CI: 35.34, 54.04) among studies their sample size were greater than or equals to 250. Those clients who have position change during hospitalization were 82% less likely to develop pressure ulcer [(OR: 0.18 (95% CI (0.07, 0.46)) than their counter part. Conclusion overall pressure ulcer Ethiopia relatively high. Position change of the client paramount the Therefore, policymakers (federal of could give special the magnitude and improve the quality of Further, meta-analysis

Those clients who have position change during hospitalization were 82% less likely to develop pressure ulcer [(OR: 0.18 (95% CI (0.07, 0.46)) than their counter part.
Conclusion The overall prevalence of pressure ulcer in Ethiopia was relatively high. Position change of the client during hospitalization had paramount benefit to reduce the burden of pressure ulcer. Therefore, policymakers (federal minister of health) could give special attention to minimize the magnitude of pressure ulcer, and to improve the overall quality of healthcare service. Further, metaanalysis study could be conducted to identify individual and health care service related factors to the occurrence of pressure ulcer.

Background
Pressure ulcers (PU) affect millions of people worldwide and nearly always occur over bony areas of the body where pressure and tissue distortion is greatest. Pressure ulcer has variation in size and 3 severity of damages to the skin, underlying tissue, muscle and over a bony prominence [1,2].
Globally, recognized as one of the five most causes of harm to clients [3]. Though pressure injury is largely preventable patient safety problem, which have a major impact on the health care system and a trivial problem to patients, relatives and caregivers [4].Likewise, pressure ulcers highly threatens the well-being of clients by increased mortality rates, decrease quality of life, longer hospital stay and increase costs for patient care [5,6].Moreover, pressure ulcer has detrimental impact on pain/suffering, disturbance of body image, delayed healing and have a negative effect on patients' overall performance [4,7].
In addition to its impact on individual and heath care system, pressure ulcers carry a significant economic burden. Pressure ulcer is considered more expensive to treat than to prevent pressure ulcers [8]. It has been estimated that the cost of treating pressure ulcer is 2.5 times higher than the cost of preventing [9]. The total annual cost for treatment of PUs in the United Kingdom being ($1.4-2.1 billion), making up 4% of the annual national health service budget [10].
A meta-analysis study revealed that the overall global prevalence of PUs using point prevalence was 14.8% [11]. Likewise, a systematic review carried out in acute care settings showed between 6% and 18.5% [12]. Other, meta-analysis report on the incidence of PU in the emergency department was 6.31% [13]. Knowing the prevalence rates of PUs are serve as basic tools to identify the severity of problem, to design preventive strategies and for efficient use of healthcare resources [14]. In addition, it would be baseline data for quality indicators to measure health care delivery within the clinical  30,33,34], fecal and urinary incontinence [33,35], worse Braden scores [6,27,33,[36][37][38], comorbidity [6,27], and nutrition[28, 34,37,39,40] was found to be statically significant risk factor for PU. However, repositioning would reduce the magnitude of pressure over vulnerable areas of the body [1].
Identification of associated factors is the primary goal to decrease the incidence of PUs [41].Hence, determining risk factors used as benchmarks to design appropriate prevention measure, to improve client safety and efficient utilization of resources [42]. Moreover, preventive measures are generally divided into four main areas: assessment of pressure ulcer development risk, skin care and initial treatment, use of pressure-reducing support surfaces and education [43].Therefore, early detection of patients who are susceptible to pressure ulcer is crucial, and it is recommended that the first skin assessment should be performed within 8 hours of hospital admission [44].
Although the attention on PU prevention has low in Ethiopia. A pressure ulcer remains pressing problem and is a major issue in nursing care. Prevention of pressure ulcers is the key role of the nurse and it is one of the quality indicator of nursing care [43]. Despite, extensive data on developed country, there is no comprehensive PU prevalence report that would serve as baseline information to improve patient safety and provide quality nursing care. Therefore, the present meta-analysis is aimed to assess the effect of position change on pressure ulcer among adult hospitalized clients in Ethiopia. Finding from the current study would serve as benchmark for policy-makers to implement appropriate preventive measure and to alleviate the pressing problem of pressure ulcer. In addition, for clinicians estimating magnitude of pressure ulcer would reflect overall quality indicator for facilities and a way to assess the efficiency of prevention strategies. Furthermore, the results of this study could serve as an input for further PU studies in Ethiopia.

Design and search strategy
This systematic review and meta-analysis was carried out by using the Preferred Reporting Items for 5 Systematic Reviews and Meta-Analyses (PRISMA) Guidelines [45]. The search strategy was developed using Population Exposure Controls and Outcome (PECO) searching guide. A two-step search strategy was used to identify all relevant literature. First, seven electronic databases were systematically searched, such as MEDLINE (via PubMed), Google Scholar, Africa journal of online, Scopus, Webscience, WHO afro library, and PsycINFO to identify relevant studies electronically.Second, a hand search of gray literature and other related articles in order to identify additional relevant research, which may use as input in our meta-analysis and the reference lists of all retrieved articles was carried out to identify additional studies. In addition, all electronic sources of information were searched the study which is done from 1 st January/ 2000 to 1 st June, / 2019. The search was conducted using the following MeSH and free-text terms : "pressure ulcer", "pressure injury", "decubitus ulcer", "bed sore", "position change" and "Ethiopia". Finally a completed PRISMA checklist has been included (Additional file 1: Table S1).

PECO guide Population
All adult clients admitted to healthcare settings with age greater than 18 years old.

Exposure
The presence of position change of the client during hospital stay.

Comparison
Absence of position change of the client during hospital stay Outcome Pressure ulcer.

Exclusion criteria
Those articles which didn't fully accessed at the time of our search process were excluded. Patients admitted with pressure ulcer. If we unable to assess the quality of each article in the absence of their full texts. Likewise, articles in which outcomes are not well defined were also excluded. Studies with poor quality as per stated criteria were also excluded from the review. Finally two authors (W.S. and Y.A.) independently evaluated the eligibility of all retrieved studies, and any disagreement and inconsistencies were resolved by discussion and consensus with the third author (T.Y.).

Outcome measurement
This review has considered studies that include all stages of pressure ulcer or equivalent as the outcome measure. Pressure ulcer is defined as a lesion of skin or underlying tissues by direct unrelieved pressure on the skin. Similarly, according to the National Pressure Ulcer Advisory Panel (NPUAP) has defined prevalence as "a cross-sectional count of the number of cases at a specific point in time, or the number of people with pressure ulcers who exist in a patient population at a given point in time" [46].

Data extraction
Data were extracted by three authors using a pre-piloted and standardized data extraction format prepared in a Microsoft excel. The data extraction sheet was piloted on 5 randomly selected papers and modified accordingly. This form was include the study characteristics, like author/s name, year of publication, study health institution, study design, sample size, prevalence, and the quality score of each study were extracted from each included article by three independent authors. Any disagreements at the time of data abstraction were reconciled by discussion and consensus.

7
The qualities of each studies were assessed using a standardized tool helps to classify risk of bias which can help to explain variation in the results of included studies. Methodological and other quality of each article was assessed by both authors based on a modified version of the Newcastle-Ottawa scale for cross-sectional study [47], a validated tool for assessing risk of bias in observational studies.
After reviewing different literatures, we declared that articles scored ≥6 points out of 10 points modified NOS components were considered to be high-quality: selection (5 points score), comparability (2 point score), and outcome assessment (3 point score). Furthermore, quality assurance check was independently performed by three authors.
Additional file 2: Table S1. Methodological quality assessment of cross-sectional studies using modified Newcastle -Ottawa Scale (NOS).

Statistical analysis
Data were abstracted by using Microsoft Excel sheet, then further analysis was done using STATA version 14 statistical software [48]. Results of the meta-analysis were reported as pooled prevalence of pressure injuries with 95% confidence intervals (CIs), p-values <0.05 were considered statistically significant. Heterogeneity across the studies were evaluated using the I 2 statistics (0% ~ 100%) [49]. I 2 statistic with a value above 75% was interpreted as representing high heterogeneity. The analysis was done using the random effects model to calculate pooled effect estimates [50]. To minimize the random variations between the point estimates of the primary study, subgroup analysis was done based on study sample size. In addition, to identify the possible sources of heterogeneity, metaregression was deployed by considering the year of publication and sample size as covariate [51]. The funnel plot graphically checked the existence of publication bias in meta-analyses. In addition, egger regression test was reported with a p < 0.05 being considered statistically significant, in order to avoid limitations of the funnel plot [52,53]. Moreover, we performed a sensitivity analysis to describe whether the pooled effect size was influenced by individual studies [54].

Search results
We found that a total of 401 articles based on systematically international database search, of these, 8 392 studies were found from seven international databases and the remaining 9 were manual search.

Study baseline characteristics
A total of 7 studies with 1,881 participants were included in this meta-analysis. Among 7 studies three were conducted in Amharic region [30,40,55], whereas the remaining was conducted in (SNNP [56], Harari [57] and Oromia [38]). Concerning, sample size most, (71.4%) of the studies, their sample size were less than 250. Based on modified Newcastle Ottawa quality score assessment almost all seven article fulfil the required quality which is 7 and above. Moreover, all studies were also a crosssectional that was conducted among clients admitted in different clinical setting of Ethiopia.

Prevalence of pressure ulcer in Ethiopia
According to the present meta-analysis evidence the pooled prevalence of pressure ulcer in Ethiopia was 11.7% (95% CI: 7.18-16.13) ( Figure 2). Using random effects model statistically significant level of heterogeneity was observed (I 2 = 90.3%; p < 0.001). The presence of significant heterogeneity among the primary studies requires the need to conduct subgroup analysis. As a result, in order to identifying the sources of heterogeneity we had deployed sub group analysis by using study sample size to determine the pooled prevalence of pressure ulcer (Figure 3). The finding of subgroup analysis, showed that, the highest burden of pressure ulcer were observed among studies groups whose sample size was greater than or equals to 250 which is 15.89% (95% CI: 13.32,18.46), I 2 = 0.0%).

Meta-regression analysis
As the test statistic shows that there were a significant heterogeneity within and between the included studies (I 2 statistics = 90.3%).Hence, in order to minimize the source of heterogeneity between the point estimates of the primary study, subgroup analysis was done based on study sample size. In addition, in order to identify the possible source of heterogeneity, we have performed meta-regression by using publication year and sample size as continuous variable of each articles as covariate of interest. Moreover, the result of the meta-regression analysis revealed that publication year and sample size were not statistically significant for the presence of heterogeneity (Table 2).

Publication bias
To identify the presence or absence of publication bias funnel plot, egger's regression test was performed. In this meta-analysis funnel plots indicated evidence of publication bias. Each point in funnel plots represents a separate study and asymmetrical distribution is evidence of the existence of publication bias and visual inspection of the funnel plot also suggests asymmetry distribution ( Figure   4). Likewise, the result of Egger's test was statistically significant for the presence of publication bias (P = 0.036). In addition, to reducing and adjusting publication bias trim and fill analysis was also performed ( Figure 5). Trim and fill analysis is a nonparametric methods for estimating the number of missing studies that might exist.

11
The finding of sensitivity analyses using random effects model revealed that no single study affected the overall magnitude of pressure ulcer (figure 6).   for the above variations might be methodological differences (i.e., data analysis and eligibility of study participants), variation in quality of care and educational preparation among health care provider, policy and strategy difference. Other plausible reasons for the observed difference between the pooled estimates could be variation in sampling, and the tool used for assessing the pressure injuries.
In the current meta-analysis, we had performed sub-group analysis based on the study sample size (i.e. sample size <250 and ≥250) in which the studies were conducted. As a result, the findings of the subgroup analysis revealed that variability was observed in the overall pooled prevalence across the category of each sample size. Among the category of sample size the highest pooled prevalence of pressure ulcer was observed from those studies their sample size were greater than or equal to 250 which revealed that 15.89%(95%CI: 13.32, 18.46).In addition the observed high heterogeneity was explored by publication year and sample size using a meta-regression analysis, although the results did not show any statistical significance. The possible explanation for this variation might be if the sample size is increase would provide the true estimate of the effect.
The current meta-analysis was also determined the association between client position change and pressure ulcer. Those who had position change during hospitalization was reduced the chance of pressure ulcer by 82% than those who had not implement position change during hospital stay. This finding is in line with the report in revised national pressure ulcer advisory panel [1] and other different study which is conducted across the globe {32-34].This work is depending on previous research, through adapted methods, a comprehensive search of the literature, and strict eligibility criteria applied. We also carry out quality assessment, with consideration of study attrition, participation and confounding factors as a means of highlighting study bias and limitations.
The current study has implications for clinical practice. The finding would serve as base line for health care provider to establishing robust preventive measures for averting pressure ulcer. The finding serves as alarming the nursing education institution to facilitate and encourage about the prevention strategy of pressure ulcer, and giving a focus on the application of standardized care. Health care setting should be adequately staffed with nurses to prevent pressure ulcer and to implement client reposition during their hospitalization. Furthermore, design different strategies and training could be deployed to enhance the implementation of client reposition and to minimize the burden of pressure ulcer as a standardize plan of care.
Even though, there were some limitations, the present meta-analysis has provided important information and synthesis evidence regarding the effect of position change on pressure ulcer; first limitation is, the present study was include only English articles were considered to conduct this nationally based review. Second, it was challenging to synthesis some additional factors meanwhile 13 they were not examined in a related approach across the studies. Third, many of the included studies did not report baseline sociodemographic characteristics of the participants. This prevented a subgroup analyses to estimate the prevalence of pressure ulcer using each variable. Furthermore, all included studies reported hospital-based populations and so this review does not consider homedwelling people with pressure ulcer. Finally, almost all included studies were cross-sectional which might weaken the strength of evidence and hinder causality inference.

Conclusion
The overall pooled prevalence of pressure ulcer in Ethiopia was relatively high. Position change of the client during hospitalization had paramount benefit to reduce the burden of pressure ulcer. Therefore, policymakers could give special attention to minimize the magnitude of pressure ulcer in order to improve the overall quality of healthcare service. Further, meta-analysis study will be needed to identify individual and health care service related factors to the occurrence of pressure ulcer.
Providing this information will inform effective strategies for preventing pressure ulcer in future.
Abbreviations CI: Confidence Interval, OR: Odds Ratio, PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analyses, SNNP: Southern Nations, Nationalities, and Peoples.

Ethics approval and consent to participate
Not applicable.

Consent for publication
Not applicable.

Availability of data and materials
The data analyzed during the current meta-analysis is available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
Not applicable.