This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Diabetes, is properly cited. The complete bibliographic information, a link to the original publication on http://diabetes.jmir.org/, as well as this copyright and license information must be included.
Chronic wounds such as diabetic foot ulcers, venous leg ulcers, and pressure ulcers are a massive burden to health care facilities. Many randomized controlled trials on different wound care elements have been conducted and published in the Cochrane Library, all of which have only a low evidential basis. Thus, health care professionals are forced to rely on their own experience when making decisions regarding wound care. To progress from experience-based practice to evidence-based wound care practice, clinical decision support systems (CDSS) that help health care providers with decision-making in a clinical workflow have been developed. These systems have proven useful in many areas of the health care sector, partly because they have increased the quality of care, and partially because they have generated a solid basis for evidence-based practice. However, no systematic reviews focus on CDSS within the field of wound care to chronic wounds.
The aims of this systematic literature review are (1) to identify models used in CDSS that support health care professionals treating chronic wounds, and (2) to classify each clinical decision support model according to selected variables and to create an overview.
A systematic review was conducted using 6 databases. This systematic literature review follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement for systematic reviews. The search strategy consisted of three facets, respectively: Facet 1 (Algorithm), Facet 2 (Wound care) and Facet 3 (Clinical decision support system). Studies based on acute wounds or trauma were excluded. Similarly, studies that presented guidelines, protocols and instructions were excluded, since they do not require progression along an active chain of reasoning from the clinicians, just their focus. Finally, studies were excluded if they had not undergone a peer review process. The following aspects were extracted from each article: authors, year, country, the sample size of data and variables describing the type of clinical decision support models. The decision support models were classified in 2 ways: quantitative decision support models, and qualitative decision support models.
The final number of studies included in the systematic literature review was 10. These clinical decision support models included 4/10 (40%) quantitative decision support models and 6/10 (60%) qualitative decision support models. The earliest article was published in 2007, and the most recent was from 2015.
The clinical decision support models were targeted at a variety of different types of chronic wounds. The degree of accessibility of the inference engines varied. Quantitative models served as the engine and were invisible to the health care professionals, while qualitative models required interaction with the user.
Chronic wounds such as diabetic foot ulcers, venous leg ulcers, and pressure ulcers are a massive burden on the health care facility [
Conventional care of chronic wounds includes [
In an attempt to increase the evidential basis and help nurses proceed from experience-based practice to evidence-based practice, several alternatives have been considered. One of these involves the development and application of health information technology systems [
As it can be seen in
Over the decades, CDSS have proven to be useful in many areas within the health care sector [
Clinical decision support models can be grouped according to different classifications. Included here are examples of the different approaches related to each classification.
The aims of this systematic literature review are (1) to identify models used in CDSS from the past decade that support health care professionals treating chronic wounds, (2) to classify each clinical decision support model, and (3) to create an overview.
The present systematic literature review follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement for systematic reviews [
Publications from MEDLINE/PubMed, Cumulative Index of Nursing and Allied Health Literature (CINAHL), The Cochrane Library, Excerpta Medica dataBASE (EMBASE), Scopus and Web of Science were searched in March 2017 to identify articles that described and discussed clinical decision support models supporting health care professionals treating chronic wounds.
The selected databases used different terminology in indexing articles, meaning that there is a risk that not all relevant articles are identified during a search. In an attempt to include all relevant articles, we used thesauruses, a systematic record in databases of subject headings used to index articles. To organize the search systematically, we grouped the search terms around three facets: ‘algorithm’, ‘wound care’, and ‘clinical decision support systems’. Further elaboration of the search terms used for eligible articles in the three facets can be seen in
The focus of this study was on models used in CDSS within the area of wound care related to chronic wounds. We, the author and the co-authors, were interested in studies that presented algorithms, models, and that were relevant for wound care as well as studies that presented wound care decision support systems or clinical decision support models. Since wound care differs depending on whether the wound is acute or chronic, we excluded studies based on acute wounds or trauma.
We excluded studies that presented guidelines, protocols, and instructions focusing on wound care since they do not require progression along an active chain of reasoning from the clinicians, just their focus. Furthermore, clinical decision support models serve as learning tools, which was of interest to us. Finally, studies were excluded if they had not undergone a peer review process.
The reference management software program Legacy Refworks (version 2.0, 2010) was used to handle the articles. In order to remove duplicates in the identified references, the functions ‘Exact Duplicates’ and ‘Close Duplicates’ were applied.
Titles and abstracts of the identified citations were read to screen the articles using the inclusion and exclusion criteria described in the previous section. The articles that remained were then read in full to extract relevant information. Afterward, cross-referencing techniques were applied on the reference lists of the included articles to identify literature that had not been discovered through the systematic literature search. The extracted information included authors, year, country, a sample size of the data, and variables describing the type of clinical decision support model [
The three facets below, shows the search strategy applied in the systematic literature review. Each facet consists of MeSH terms and synonyms. Between each MeSH term and synonym, the Boolean operator OR is used and between each facet the Boolean operator AND is applied.
Facet 1 (Algorithm) | Facet 2 (Wound care) | Facet 3 (Clinical decision support system) | ||
Regression analysis OR Statistical models OR Linear models OR Loglinear model OR Multivariate logistic regression OR Logistic models OR Regression analysis OR Logistic regression OR Artificial neural network OR Theoretical model OR Computer simulation OR Prediction OR Bayes theorem OR Prognosis OR Forecasting OR Artificial intelligence OR Artificial intelligence OR Algorithm-based OR Model-based OR Model OR Algorithms OR Prescriptive OR Pattern recognition OR Data mapping OR Text mining OR Data mining | AND | Therapy OR Wound treatment OR Wound management OR Wound assessment OR Pressure ulcer care OR Wound care OR Skin care OR Skin care OR Foot care OR Larval therapy OR Autolytic debridement OR Chemical debridement OR Mechanical debridement OR Surgical debridement OR Debridement | AND | Clinical decision support systems |
To reduce bias during the selection and reviewing process, the author, together with one of the co-authors, systematically went through each article, discussed the scope of each article and decided whether an article was relevant in proportion to our systematic literature review. The interrater reliability was not calculated in this study. However, it could have been prudent.
The included models used in the CDSS from the studies were subsequently described and classified according to selected variables, as defined by van Bemmel and Musen [
The flowchart visualises the selection process of the articles included in the systematic literature review.
A total of 4 of the 10 (40%) articles from the systematic literature review presented a quantitative decision support model [
The applied techniques in these 4 quantitative decision support models covered the following: Neural Networks, Support Vector Machines, Random Forest Decision Tree, Bayesian Belief Networks and Logistic Regression (
The professionals who had access to 3 of the 4 (75%) quantitative decision support models were health care professionals who detect, estimate, diagnose, and register crucial tissue measurements for pressure ulcer diagnosis. They are also health care professionals who undertake wound care for chronic wounds (
The accessibility of the inference engines of the 4 quantitative decision support models was low. The 4 quantitative decision support models should help health care professionals in decision-making situations, but health care professionals could not follow the statistical processes performed on the data material by personal inspection. They could only see the outcome of the statistical processes (
The focus in each of the 4 quantitative decision support models varied. The quantitative decision support model by Veredas et al [
An overview of who the publication authors were, the year the publication was published and where the publication was published.
Reference No. | Publication Authors | Year Published | Country Where Published |
[ |
Veredas FJ, Luque-Baena RM, Martín-Santos FJ, Morilla-Herrera JC, Morente L | 2015 | Spain |
[ |
Forsberg JA, Potter BK, Wagner MB, Vickers A, Dente CJ, Kirk AD, Elster EA | 2015 | US |
[ |
Mukhejerjee R, Manohar DD, Das DK, Achar A, Mitra A, Chakraborty C | 2014 | India |
[ |
Veredas F, Mesa H, Morente L | 2010 | Spain |
The table provides an overview of which type and size of data the models were based on, and the applied techniques in the clinical decision support systems.
Reference No. | Data Presented in the Article | Applied techniques in the clinical decision support systems |
[ |
Data consisted of (n=113) images of pressure ulcers on sacrum and hips. | K-means clustering algorithm for image segmentation. Three machine learning approaches (1) Neural Networks, (2) Support Vector Machines, and (3) Random Forest Decision Trees |
[ |
Data consisted of (n=73) participants (a mix of soldiers and civilians) with at least one extremity wound >75cm2. | Parametric statistical and machine learning methodologies (1) Bayesian Belief Networks, (2) Random Forest Analysis, and (3) Logistic regression using Least Absolute Shrinkage and Selection Operator. Statistical differences between the continuous variables and wound outcomes were evaluated using the Mann-Whitney U test and the post hoc Tukey-Kramer assessment. |
[ |
Data consisted of (n=74) images of chronic wounds from the Medetec medical image database. | Fuzzy divergence-based thresholds used for wound contour segmentation. For wound tissue classification (1) Bayesian classification, and (2) Support vector machine. |
[ |
Data consisted of (n=113) images of sacrum and hip pressure ulcers. | Image processing techniques: filtering, kernel smoothing by the mean shift procedure and region growing. Statistical analysis: (1) A hybrid approach based on Neural networks, and (2) Bayesian classifiers. |
An overview of the quantitative decision support models’ accessibility of the inference engines, what type of wounds it focuses on and the type of professionals who have access.
Reference No. | Accessibility of the inference engines of the system | Type of Wounds | Professionals, who have access to the system |
[ |
The clinical decision support model aims to help clinicians in decision-making situations. Health care professionals cannot access the inference engine and cannot follow the statistical processes performed on the data by personal inspection. They can only see the outcomes of the statistical processes. | Pressure ulcers | Health care professionals who detect, estimate, diagnose and register important tissue measurements for pressure ulcer diagnosis |
[ |
The clinical decision support model aims to improve decision-making when surgeons need to know if they must close or cover a wound. |
Chronic wounds | Surgeons in hospital settings |
[ |
The decision support model helps health care professionals identify necrotic tissue within chronic wounds. Clinicians cannot access the inference engine. They can only see the outcomes of the statistical processes. | Chronic wounds | Health care professionals who undertake wound care for chronic wounds |
[ |
The decision support model helps health care professionals care for pressure ulcers. The health care professionals cannot access the inference engine and cannot follow statistical processes. They can only see the outcomes of the statistical processes. | Pressure ulcers | Health care professionals who detect, estimate, diagnose and register important tissue measurements for pressure ulcer diagnosis |
The quantitative decision support model by Mukherjee et al [
Pressure ulcers were the focus of 2 of the studies [
There were 6 articles (6/10, 60%)from the systematic literature review that presented a qualitative decision support model [
The applied techniques in the 6 qualitative decision support models included Meta-Analyses, Systematic Reviews, Literature Reviews, Expert Face Validations, Answers from Questionnaires, Expert Panel Discussions, Task Force of Clinical Experts and Consensus Panels (
The professionals who had access to 3 of the 6 (50%) qualitative decision support models were health care professionals who undertake wound care for chronic wounds (
An overview of who the publication authors were, the year the publication was published and where the publication was published.
Reference No. | Publication Authors | Year Published | Country Where Published |
[ |
Alvey B, Hennen N, Heard H | 2012 | US |
[ |
Beitz JM, van Rijswijk L | 2012 | US |
[ |
Smith G, Gibson E | 2013 | Great Britain |
[ |
Kravitz SR, McGuire JB, Sharma S | 2007 | US |
[ |
LeBlanc K, Baranoski S, Christensen D, Langemo D, Sammon MA, Edwards K, Holloway S, Gloeckner M, Williams A, Sibbald RG, Regan M | 2013 | US |
[ |
McNichol L, Watts C, Mackey D, Beitz JM, Gray M | 2015 | US |
The accessibility of the inference engines of the 6 qualitative decision support models was high. Health care professionals could follow the decision process from start to finish; furthermore, they were required to take an active part in assessing the wounds.
The focus of each of the 6 qualitative decision support models varied. In the qualitative decision support model of Alvey et al, [
Diabetic foot ulcer surgery was the focus of 1/6 (17%) studies [
Recent studies have shown that it can be valuable to apply CDSS in clinical settings to increase the quality of care and generate solid bases for evidence-based practice [
Chronic wounds are complicated to treat and challenging for health care professionals. In decision-making situations where health care professionals need advice related to chronic wound treatments, health care professionals are, according to the literature, often forced to rely on their own experience rather than having evidence available that they can follow [
The optimal type of decision support model may not be easy to identify. The techniques applied in the clinical decision support models reviewed here were classified into two groups: 4 studies were based on quantitative models, and 6 studies were based on qualitative models. One of the strengths of the quantitative decision support model classification is that outcomes of statistical processes are objective, reproducible and in accordance with the clinical data [
It is commonly accepted that accessibility of the inference engine is important [
One might argue that IT systems should be tailored to specific professional groups to address the needs of nurses or doctors optimally, for example [
Several of the existing checklists are addressed to conventional study designs such as randomized controlled trials, cohort studies and qualitative semi-structured interviews [
It is a complicated task to build a search strategy which reflects an accurate inventory of what has been done within a research area. Primarily because hits often reflect the conducted search rather than reflecting the reality within a research area. However, there are several initiatives to avoid this problem. In our study, our preliminary search strategy consisted of 4 facets, respectively: Facet 1 (Algorithm), Facet 2 (Wound care), Facet 3 (Clinical decision support system) and Facet 4 (Wound). Facet4 consisted of a mix of MeSH terms and synonyms and these were: “foot ulcer,” “diabetic foot,” “skin ulcer,” “leg ulcer,” “decubitus,” “chronic wound,” “venous foot,” “venous ulcer,” “pressure ulcer,” “ulcer,” “wounds and injuries,” “varicose ulcer,” “ulcer wound,” and “diabetic foot”. Between each of the search terms in Facet 4 the Boolean operator “OR” was inserted to achieve as many hits as possible. When running the search with the 4 facets, we did not receive any hits, maybe because it was too narrow. Hence, we refined our search strategy and made it broader. Subsequently, we removed search query facet 4, so the search strategy instead only consisted of Facet1, Facet2 and Facet3. When running the revised search strategy, we retrieved several hits as depicted in our flowchart (
We used the classification suggested by Bemmel and Musen [
There were 10 clinical decision support models identified. Of these, 4 (40%) were quantitative decision models and 6 (60%) were qualitative decision support models.
Three (3/4, 75%) of the quantitative decision support models were applicable for all health care professionals who detect, estimate, diagnose and register essential tissue measurements for pressure ulcer diagnosis or who undertake wound care for chronic wounds. The fourth (1/4, 25%) quantitative decision support model was applicable for surgeons who work in a hospital setting. Two qualitative decision support models were designed specifically for registered nurses, licensed practical nurses and specialized nurses such as wound care nurses who provide critical support for tissue viability services. One qualitative decision support model applicable for surgeons who undertake diabetic foot ulcer surgery.
The degree of accessibility of the inference engines varied. The 4 quantitative decision support models served as engines and were invisible to health care professionals. The 6 qualitative decision support models required interaction with health care professionals.
The clinical decision support models were targeted towards different types of chronic wounds. Two (2/4, 50%) of the quantitative decision support models focused on pressure ulcers, while the remaining 2/4 (50%) focused on chronic wounds. One of the 6 (17%) qualitative decision support models explicitly focused on diabetic foot ulcer surgery, 4 (4/6, 67%) focused on chronic wounds, and the last (1/6, 17%) qualitative decision support model focused on pressure ulcers.
More research is needed to develop clinical decision support models targeted at health professionals treating chronic wounds. Given the growing focus on evidence-based care and the availability of increasing amounts of data, the arguments for a quantitative approach to decision models in future work are strengthened. Future research should also address problems with accessibility in quantitative systems by developing methods for augmenting users’ understanding of the processes in the quantitative models.
The table provides an overview of which type and size of data the models were based on, and the applied techniques in the clinical decision support systems.
An overview of the qualitative decision support models’ accessibility of the inference engines, what type of wounds it focuses on and the type of professionals who have access.
Clinical Decision Support Systems
Cumulative Index of Nursing and Allied Health Literature
Consolidated Standards of Reporting Trials
Excerpta Medica Database
Medical Literature Analysis and Retrieval System Online
Medical Subject Headings
Preferred Reporting Items for Systematic Reviews and Meta-Analyses
Prospective Review Database
This systematic review study received no financial support for the research, authorship, and publication of this study.
None declared.