Introduction
It is hard to disagree that the purpose of every medical care worker is to help patients and make their experience in a healthcare facility positive. Consequently, one of the issues healthcare professionals still have to address is the number of clinical errors that reduce the general performance and harm patients. One of the recent implementations that makes it easier for clinicians to define the necessary treatment is computerized clinical decision support systems (CDSSs). Such systems serve as helpful tools for both patients and medical workers and make their collaboration more informed. At the same time, not all aspects of CDSSs are studied, and researchers are yet to prove whether such systems are important to a major extent. The purpose of this paper is to analyze and discuss the 2018 article “Effects of Computerized Decision Support System Implementations on Patient Outcomes in Inpatient Care: A Systematic Review” by Varghese et al.1
Aspect of CDS
It is possible to say that the article is of significant importance for healthcare professionals, who indeed need to know the topic discussed in the paper. Varghese et al. strive to answer two crucial research questions: “are there specific disease entities for which CDSSs succeeded more frequently?” and “to what extent are the reported outcome effects clinically important?”1(p594) Therefore, the results of their study may be of extra value for medical workers who already work with CDSSs or are considering such an option.
In their research, Varghese et al. focus on the significance and effectiveness of CDSSs. The aspect they explore is whether the use of CDSSs can improve patient outcomes and whether such improvements are actually valuable for the healthcare practice. What is more, another aspect that Varghese et al. discuss is the appearance of unique effects of these computerized systems on a number of particular diseases.
Study Methodology
In order to answer the research questions and gather a synthesized volume of relevant information, the authors conducted a systematic literature review. According to Varghese et al., “the strategy was adapted to focus on terms relevant for CDSS and patient outcome assessments,” and relevant studies were searched in Cochrane Reviews, Cochrane Trials, and Medline.1(p594) However, not all articles on this topic could appear insightful and informational, and it was essential for the authors to focus on research published between 2005 and 2016. After a number of selection stages, out of 20,835 articles, the researchers finally chose a total of 70 most relevant and insightful studies to focus on them. Further, “for each study, 2 physicians independently mapped patient outcome effects to a predefined medical effect score” in order to evaluate the reported outcome effects’ clinical influence.1(p596) After that, the findings have been systematized and analyzed to find differences and similarities between the discoveries and conclusions. This methodology allowed the authors to get an understanding of the available findings on the topic, as well as determine literature gaps.
Endpoints and Outcomes Studied
The reason for the authors to conduct their research was the mixed findings of other researchers and the limited information regarding the issue. For instance, Varghese et al. state that the most relevant meta-analyses of the impacts of CDSS on patient outcomes “are limited to randomized controlled studies.”1(p594) This is probably due to issues that are methodological in nature, including potentially stronger reporting biases and false precision.1 Therefore, the authors find it essential to identify whether there are any particular disease entities for which the computerized systems demonstrate more frequent success. If proved positive, this endpoint can contribute to the general consideration of the value of CDSSs, as well as their more extended utilization in healthcare facilities. Another endpoint Varghese et al. try to research is the extent to which the reported outcome effects are clinically significant.1 This endpoint is essential because if the impact of CDSSs is not strong enough, then other solutions to improve patient outcomes should be found.
Positive and Negative Aspects Identified
The investigation conducted by the authors of the article allows them to identify specific positive and negative aspects of CDSSs. For example, the latter includes the fact that these systems need to be tested and validated carefully before the implementation, and all clinicians have to be skillful and trained to utilize these systems. Otherwise, there is a severe risk for patients if an unprepared medical worker uses a CDSS.1 Another negative aspect is that the results are inconsistent, and it is impossible to predict whether the impact of CDSS in a particular case will be successful. Consequently, one may still doubt the necessity and benefit of using such computerized systems. As for the positive aspects, it is possible to say that CDSSs, in general, actually make the clinicians’ decision-making process easier and more informed.1 What is more, the systems usually have valuable effects on patient outcomes, reducing the potential harmful impacts.
Results
It is possible to say that the results of the study are rather contradicting. Overall, out of the 70 selected articles, “five (7%) reported reduced mortality, 16 (23%) reduced life-threatening events, and 28 (40%) reduced non–life-threatening events.”1(p596) What is more, “20 (29%) had no significant impact on patient outcomes, and 1 showed a negative effect.”1(p597) It is essential to mention that the researchers managed to answer their first research question by identifying a total of six disease entities in which the use of CDSSs has the most beneficial effects. These entities are “blood glucose management, VTE prophylaxis, blood transfusion management, acute kidney injury prevention, physiologic deterioration prevention, and pressure ulcer prevention.”1(p597) Therefore, when using the computerized systems for these entities, it is more likely for clinicians to prevent significantly harmful events and achieve improved patient outcomes.
The Need for Further Investigation
Overall, as stated by Varghese et al., there are several gaps in the literature, which leads to contradicting conclusions and a lack of evidence. What is more, various biases and limitations make it challenging to rely on the investigations’ results. Consequently, Varghese et al. recommend that the following RCTs related to this topic should carefully consider the lack of confounders adjustment, contamination effects, and the size of their sample.1(p598) What is more, it is required that future studies can identify a disease-independent common rule that explains why for some disease entities, the computerized systems succeed more. As for defining the extent to which the CDSSs effects are crucial, further researchers need to consider this as their primary question. Further, in the previous paragraph, it is mentioned that a negative effect of computerized systems had been identified in one of the selected studies. Thus, such an issue needs to be researched further to exclude any possible harmful impacts of these CDSSs.
Conclusions and the Effect of the Article on the Healthcare Practice
To draw a conclusion, one may say that the article is quite significant for the clinician practice and contributes greatly to the knowledge about CDSSs. The authors conclude that, although about seventy percent of the selected papers report improved patient outcomes, major differences related to the clinical importance of these outcomes still exist.1(p601) As for the identified disease entities, this discovery is of extreme value for the healthcare practice. While being aware of the list of the six entities, clinicians and healthcare facilities that still underutilize computer-assisted decision-making systems can now give them special consideration.
Reference List
1. Varghese, J., Kleine, M., Gessner, SI., Sandmann, S., Dugas, M. Effects of computerized decision support system implementations on patient outcomes in inpatient care: a systematic review. J Am Med Inform Assoc. 2018; 25 (5): 593-602.