The author of this paper IS a nurse educator working in an Intensive Care Unit (ICC) within the north-west region of England. For the purpose of this paper, the author has chosen a systematic review (SIR) and a narrative review (NOR) and alms to critically analyses these reviews using the Critical Appraisal Skills Programmer (CAPS) tool. The CAPS tool has been developed by the Solutions for Public Health (SSP) (formerly known as Public Health Resources unit), which Is controlled by the National Health Service (INS) (SSP, 2006).
The tool Is based on the guidelines formed by a group of clansman at Master university, Canada (SSP, 2010). Additionally, the author will apply the Preferred Reporting Items for Systematic Reviews and meta-analyses (PRISMS) 2009 checklist as a gulled to support this discussion. The PRISMS checklist is a 27 item checklist, subject to modification as new evidence emerges (Liberate et al, 2009). Accordingly, this critical appraisal will rigorously analyses the strengths and weaknesses of the papers chosen for this discussion and assess its relevance to practice. Medication errors and adverse drug events vary from 5. O 87. 5 incidents per 1000 patient days in CICS (Wilier et al, 2010). Computerized Physician Order Entry (COPE) was introduced to minimize these errors and promote patient safety within the ICC setting (Causal et al. 2003; Rothschild, 2004). The SIR and NOR that the author has chosen highlight the importance of COPE systems in promoting patient safety and reducing avoidable harm. The papers can be accessed using these references: Causal R, Shania KEG and Bates W (2003) ‘Effects of computerized physician order entry and clinical decision support systems on medication safety: a systematic review’ Archives of
Internal Medicine 1 63 up. 1409-1416 and Rothschild J (2004) ‘Computerized physician order entry in the critical care and general inpatient setting: A narrative review’ Journal of Critical Care 19 (4) up. 271-278. Moreover, the author will conduct these appraisals alongside one another, similar to a constant comparative analytical process. Bastion et al (2010) acknowledge that there are 75 trials and 1 1 SIR published each day In the world of healthcare research. However, It has also been recognized that only a small minority of the published reviews are SIR and majority of the reviews are non-systematic NOR (Bastion et al, 2010).
So what Is the difference between a SIR and NOR? SIR are reviews that are conducted using expelled strategies to critically appraise and synthesis issues on a given topic and NOR are reviews conducted using expert opinion based on the subjectivity and experience of the author writing it (Dickers, 2009: Misunderstanding and Maladapted, 2006). There is a general consensus that SIR is superior to NOR and research papers often criticism NOR for being non-systematic Pettier, 2001). Nonetheless, researchers argue that Mrs. are of equal importance (Carr-Hill, 1997).
Also, Dickers (2009) argues that both SIR and NOR are of relevance to reactive and either of them can be useful or prone to errors and flaws. Therefore, a rigorous analysis of any review has to be carried out before Judging its outcomes. As mentioned earlier, the author has chosen the CAPS tool provided by the SSP to appraise the reviews. The rationale for using this tool is because the tool was specifically designed for appraising evidence based literature in health and social care and relates knowledge into practice (SSP, 2010).
Therefore, this would help the author to ‘critically read’ the paper to be evaluated and provide the author with the ability to exclude articles that cannot be deemed relevant to practice and accept those articles with ample quality. Furthermore, it must be understood that the author is not going to appraise the entire paper using the PRISMS checklist and will use the checklist as a guide to discuss the aspects that have been omitted from the CAPS tool.
The author feels that using two different tools would increase the rigor of the appraisal and its relevance to practice. In the forthcoming sections of this essay, the author will attempt critically analyses the two papers using the CAPS tool for Sirs. Formulation of the research question Formulation of the research question guides the type of data collection for the study and states what the study will ‘hypothesis’ (Polite and Beck, 2010).
The CAPS tool acknowledges that clearly-focused research question will describe the population studied, the intervention given and the outcomes considered (SSP, 2006). The research question in the SIR is clearly formulated. The population of patients studied were general and critically ill hospital patients, the intervention was a SIR and outcome to be assessed was about the safety of the COPE and clinical decision purport systems or CADS (CADS is a computerized medication prescription system) based on hierarchy of outcome measures (kindly refer to table 2 of SIR) within ICC settings.
Likewise, in the NOR, the population of patients studied were general and critically ill patients, the analysis was done via a NOR and the purpose of the study was to assess the results of COPE on clinical and surrogate outcomes in hospitalized patients in both general and critical care settings. Based on this, the author acknowledges that the SIR has adopted a narrow and more specific research question as compared to the NOR.
This is because the SIR aims at assessing the effects of COPE on CADS in critically ill patients and the NOR assessed effects of COPE in ‘medication prescribing (CADS), ‘diagnostic test ordering and result management’ and ‘injury prevention and resource utilization interventions’. Therefore, the author felt that the SIR had adopted a narrow search strategy to address their research question and the NOR has adopted a wider strategy.
Nevertheless, there is no indication of any review protocol being used in both the studies and no information has been presented about formulation of research question I. E. SE of PICKS (population, intervention, comparison intervention, outcomes and study design). However, the PICKS can be understood by reading the papers in-depth. Title and abstract It is evident that both the titles of the reviews identify that the report is a SIR and NOR. Additionally, both the studies did not provide a fully structured summary in the abstract.
The PRISMS checklist necessitates provision of a structured summary interventions, study appraisal and synthesis methods, results, limitations, conclusions and implications of key findings, systematic review registration number [Liberate et al (PRISMS checklist #2), 2009]. Although the SIR provides with information on background, methods, results and conclusion; it does not provide any further information on the other aspects. On the other hand, the NOR includes a general abstract without any sub-headings and minimal information in the abstract.
Study selection for the review The SIR divided its study selection into 2 groups; group one evaluating COPE with CADS and group 2 evaluation CADS alone. There were 12 studies that met their inclusion criteria for study design and outcomes. All these studies chosen address the Sir’s question and there were 6 randomized controlled trials (Arcs), 2 Arcs with resolve design, 1 prospective (level 2) study, 2 retrospective (level 3) studies and 1 mixed study (level 2 and 3). The NOR chose 18 studies altogether, which comprised of 2 studies in critical care settings, 8 in general care settings and 8 in mixed care settings.
All the chosen studies address the research question and were classified into three groups (effects of COPE in ‘medication prescribing (CADS), ‘diagnostic test ordering and result management’ and ‘injury prevention and resource utilization interventions’). There were 4 time series studies (level 3), 6 mixed studies (level 2 and ), 1 crossover time series study and 7 Arcs. The review excluded studies which evaluated COPE without CADS. On further analysis of the studies, the author discovers that both the studies did not identify the population of patients to be involved I. . Adult or pediatric patients in their search strategy. Nevertheless, the SIR did not involve any pediatric studies and the NOR included both adult and pediatric studies. Furthermore, the SIR has identified that it included articles with a minimum level 3 (hierarchy of evidence presented in table 1 of SIR) in its study and although the NOR included articles with a minimum level 3 in its review it does not point out this feature in its study selection and identification. The study selection process I. E. Inclusion and exclusion have not been put into a flow diagram.
Nonetheless, the data extracted has been summarized with citations, headings and simple summary data for each study. Literature search strategy Both the studies performed a thorough literature search on databases, namely, MEDICINE and Cochrane review databases and identified use of keywords specific to the research question. However, the NOR extended its search to Predefine, EMBASSY and all IBM review databases. Nonetheless, the author felt that both the researchers could have extended its search to other databases, namely, OBESE, OVID and other foreign databases.
Furthermore, both the studies retrieved articles from reference lists of other literature. Information on literature search is clearly outlined and the databases searched have been listed in the reviews. Conversely, there is no information in both the studies about searches for unpublished studies and search for non-English studies. Greenshank (2010) recognizes the importance of conducting an arduous search and recommends the reviewer to use foreign language literature ND search a minimum of least 10 databases to obtain majority of the literature available to the topic.
However, when the author searched for foreign language literature using similar search strategies as explained in the papers, no articles were The NOR specifies the timeshare of the search strategy (between 1966 and 2001) and the SIR does not mention this. Additionally, the NOR had more studies included within its review probably because it had searched for articles within more databases as compared to the NOR. Therefore, the author feels that the NOR has less evidence of publication bias compared to the SIR. This issue has been discussed in more detail in the next section of this paper.
Assessment of quality of included studies Both the studies included Arcs, prospective, retrospective trials and time series studies for their review. All these studies are between levels 1 to 3 in the hierarchy of evidence. The SIR had 3 authors and they resolved their disagreements through discussions. The NOR had only 1 author and therefore retrieved articles were appraised by the author alone. Likewise, when the author of this paper made a search using similar search strategies, the same articles that have been explained in he respective papers were retrieved.
Therefore, the author would like to address that the articles for the SIR and NOR were the best available articles at that point of time. Nonetheless, for assessing the effect of COPE along with CADS, the SIR did not include the study conducted by Evans et al (1998), a prospective and retrospective study in which COPE and CADS were used to manage and order anti-infective medications in an adult ICC (n=1681 patients). Additionally, another study by Mullet et al (2001), a study similar to Evans et al (1998), but on a pediatric population of ICC attends (n= 1768 patients) was not included.
Similarly, the NOR did not include an ARC by Overhang et al (1997), which aimed in assessing the impact of physician and faculty reminders in general medicine ward of Indiana hospital to assess the effects of COPE on medication prescribing. The omission of these articles into both of these studies could have led to significant publication bias. Accordingly, Gargle (2000) recognizes that loss of studies in small numbers can result in significant bias. Therefore, the author of this paper identifies that the SIR and NOR appraised could have flawed results.
Nonetheless, the author recognizes that the results of the missed studies coincide with the findings of the SIR and NOR and as a result the articles missed from the reviews could not have been neglected deliberately. Moreover, if the researchers had included these studies within their reviews it would have given more strength to their conclusions. Results of the studies The results of the SIR have been clearly displayed and have not been combined. The tables 3 and 4 of the SIR present with the analysis of the trials and identify the effectiveness of COPE with CADS and COPE without CADS respectively.
The SIR has systematically analyses the studies and presented it in a tabular form under the following headings; source, study description, design, outcomes and results. The results of the individual studies cannot be statistically combined and meta-analyses. This is because the studies included in the review were heterogeneous. However, each trial within the study is presented with its accurate statistical data. The results of the SIR conclude that the COPE and CADS can reduce medication errors in practice, however the studies chosen lack power to detect occurrences of adverse drug events. D 3 describe the individual studies and the effectiveness of COPE with CADS (medication prescribing), diagnostic test ordering and injury prevention. The studies have been presented in tabular form under the following headings; reference, population and setting, design, description, outcomes and results. The results in the NOR cannot be combined because of the heterogeneity of studies included in the review. The NOR concludes that COPE can be beneficial with CADS to reduce medication errors; however, additional research is recommended to Judge long-range effectiveness.
Due to the heterogeneous nature of the studies included within the views it has not been possible to conduct any further meta-analyses or design forest plots with confidence intervals. Therefore, there was no combined statistical analysis of the studies done. Nevertheless, both the reviews have presented results of the individual studies with the essential statistical information of the chosen trials. Nonetheless, the author feels that the reviewers could have included an additional column for description of bias within the selected studies.
Outcomes of the reviews The SIR acknowledges that the use of COPE with CADS significantly decreases medication errors. It also recommends conducting large sample studies to successfully identify key concepts of COPE and CADS in medication prescription. It also discusses costs and benefits of COPE and CADS and potential for harm in the absence of such systems. The NOR acknowledges that the use of COPE with CADS decreases medication errors and recommends the use of such systems to reduce medication errors. The outcomes of these reviews have been discussed by the researchers from the perspectives of patients, professionals and hospital settings.
The SIR has taken safety and cost into consideration; however, the NOR has acknowledged that it was not possible to describe these issues. Both the reviews recommend larger studies to assess the effectiveness of CADS systems and focus on accepting and integrating these technologies into practice. In the next section, the author will discuss the significance of these papers into practice. The studies have summarized the main findings of the trials. They have considered its relevance to patients and medical institutions who will be primary stakeholders cited in the aims of their study.
Additionally, both the studies have discussed their limitations, forever, they fail to discuss and report bias in their respective reviews. Therefore, this might have led to selection, publication and researcher bias within both studies. Furthermore, the SIR acknowledges that Dry. Bates (one of the co-writers of the review) is a consultant to Medicals, which develops web-based CADS for radiology test ordering. On the other hand, the NOR also does not report any source of funding; however, thanks Dry. Bates for his review and support.
Overall, no other information on funding or role of fenders has been mentioned in both the reviews. This means hat the reviews were done by the authors without the influence of any software Conclusions of the reviews The SIR concludes that COPE and isolated CADS significantly reduces medication errors and recommends a need for conducting research on evaluating such systems and identifying their advantages. The NOR also reports that COPE and integrated CADS reduces medication errors and recommends additional research to assess long- range effectiveness of COPE in clinical settings.
Therefore, the aims of the study correlate to its conclusions and coincide with their respective research questions. Author’s conclusions and relevance of reviews to practice Reviews are vital to good scientific and clinical practice and understanding of current knowledge is essential to design new studies. Appraising the reviews would help the readers to understand the topic better and view it objectively. In this section, the author will be evaluating the reviews and provide an objective conclusion. The SIR has a clearly focused question with well defined interventions.
However, it fails to complete a comprehensive search strategy and does not clearly define the inclusion and exclusion criteria. A much more detailed description of inclusion and exclusion criteria, including a flow diagram of studies included and excluded would have been beneficial. Additionally, widening the search strategy could have allowed the authors to include more studies within the SIR. This would have increased the validity of their findings and added more rigor to the conclusions. The SIR does not present any data on risk of bias within the chosen studies.
Nonetheless, the review has summarized key findings and provided general interpretation of the results along with recommendations for future research. The NOR aims to address too many issues thin the research question; nonetheless, it has adopted a wider search strategy and broadly identifies inclusion and exclusion criteria. There is no flow diagram of study selection for the review and excluded studies have not been mentioned. The NOR does not present any data on the risk of bias within the chosen studies. The results of individual studies have been summarized and recommendations have been provided for future research.
The author has critiqued these articles using the CAPS tools (SSP, 2006) and PRISMS checklist (Liberate et al, 2009) and these tools were developed after these articles were published. Nevertheless, on comparing the steps of conducting a review with the then widely used checklist (the QUORUM statement), the reviews still fail to adhere to the issues mentioned above. So the next question that comes to the authors’ mind is that whether we apply these papers to practice. The author believes that neither the SIR with its confined approach nor the NOR with its uniqueness has satisfied the methodology of the review process.
Nevertheless, the NOR has been methodically comprehensive and covers multiple issues through a systematic process. Similarly, the SIR systematically analyses the need for effectiveness of the proposed intervention and additionally discusses issues of costs and potential for harm. The author feels that the SIR and NOR compared in this paper included studies with homogeneity and this would have enabled them to conduct statistical analysis, providing more rigor to the findings of the study. The author would like to acknowledge that although the reviews have flaws, they present with valuable information.
The author considers that these articles would be of relevance to practice if it were to be considered at the time of the reviews being published (year 2003 and 2004). Although they have not fully adhered to specific guidelines for designing the reviews, they provide useful information about the safety of use of COPE and CADS in general hospital settings. Based on the appraisal, the author would have proposed for a change in practice and requested for using COPE and CADS within the authors’ institution.
This is mainly because the primary issue reviewed in both papers were about patient safety related to medication prescription. In conclusion, this appraisal demonstrates that both reviews are suitable for reviewing extensive subject matter and either of them can be adapted effectively to study clinical topics. Many researchers believe that the NOR and SIR have to be used accordingly and even the most practically designed reviews are prone to errors (Dickers, 2009; Collins and Fuser, 2005; Hammerless, 2002).
There is definitely a role for both these types of reviews in research although they are subject to constant criticism. However, the main objective of conducting such reviews would be to ensure that they are explicit, comprehensible and replicated easily by readers.