A critical appraisal is a procedure of cautiously and methodically probing a study to review its trustworthiness and its value and importance in a specified context. It is an indispensable proficiency for evidence-based medicine as it allows medical practitioners to acquire and use research proof dependably and competently
Everybody would like to benefit from the most excellent medical care. To attain this, there is a need for trustworthy information about what could harm or help in making decisions health-wise. According to Cullum & Ciliska:
Research involves gathering data, then collating and analyzing it to produce meaningful information. However, not all research is good quality and many studies are biased and their results untrue. This can lead us to draw false conclusions. So, how can we tell whether a piece of research has been done properly and that the information it reports is reliable and trustworthy? How can we decide what to believe when research on the same topic comes to contradictory conclusions? This is where critical appraisal helps (Cullum & Ciliska, 2008).
Therefore this paper is going to focus on the critical appraisal of the randomized controlled trial to evaluate a nurse-led program of support and lifestyle management for patients awaiting cardiac surgery.
This program was founded on early research proposing development in risk aspects contributing to coronary ailments as patients waited for cardiac operation/surgery. The method used was based on randomized people listed for coronary-artery bypass surgery. As shown in the article by (Goodman et al, 2007); “the primary outcome measurements were anxiety, blood pressure, cholesterol, length of stay and body mass index. Costs of the intervention were also collected”.
The experiment was justifiable (results were valid) since the patients were randomized. The experiment was run as a randomized controlled trial (RCT). It was done with a cardiac home care and treatment team giving intervention measures. The patients’ perception together with that of the nurses was taken in qualitative pre-arranged interrogations and target groups being reported independently. “Costs of all in-patients, out-patients, community contacts, and the homecare contacts were analyzed to look for any impact on costs and healthcare utilization” (Goodman et al, 2007). As shown by (Goodman et al, 2007) again, “computerized random number allocation by a third party was used to allocate patients to the intervention or control group in a 1:1 ratio of 2.2.5” (Goodman et al, 2007).
On the concealment of the randomization, it is shown that to some degree the subjects and assessors were not blind to allocation. In the article, the authors indicated that “the estimates of effect size had wide confidence intervals and the randomization process was unusual in that the patients were allocated to their group before recruitment” (Goodman et al, 2007).
The patients were also analyzed into groups randomized since all of them were randomly allocated and analysis done in their clusters. The ratio at randomization was also reliable to the analysis.
The groups were shown to be similar in all known determinants of outcomes as the article does not state that the patients were harmonized before randomization. Again the groups were not reported anywhere as per the characteristics they had. On the other hand, the article does not show if results were trivial when weighed against both clusters
Indeed, the Subjects of the study were not harmonized preceding the randomization. The inclusion criteria on the other hand limited the test to patients who seemed to have identical medical traits. The clusters were shown to have the same comparability on demographic and medical characteristics gathered at baseline. The article on the other hand does not clarify whether all resulting determinants were non-considerable when evaluated in both clusters. The resulting baseline scores for the clusters were not accounted for in the article, so it could or could not be considerably dissimilar between the clusters.
In this article, the intervention and control groups retained a similar prognosis after the research was started. The patients were again aware of the group allocation as
It is evident in the article that there was no blinding in allocation. The authors do not indicate blinding them in any way although it is not verified if the patients were given comprehensive information concerning the trial and as a result alerts to the nature of the investigational and control clusters. There is the possibility that some subjects in the control cluster still thought they were part of the other cluster. The only information we have as shown by the authors is that; “the estimates of effect size had wide confidence intervals and the randomization process was unusual in that the patients were allocated to their group before recruitment” (Goodman et al, 2007).
Although it is not clear whether clinicians were aware of group allocation, the answer here is probably yes since the whole research was based around them. The clinicians are the ones with most of the information regarding such studies and they are supposed to be informed before the study/research for better results. Again the clarity of this issue is obscured by the fact that the clinicians were mostly engaged in making out potential and suitable subjects. On the other hand, it can be true that they may not have been aware of the allocation of exact patients as randomization was done after screening processes. However, nurses who were in charge of conducting the research should have been aware of the allocation. Conversely, patients could have been able to self refer through advertisements.
In this article, outcome assessors should have been aware of group allocation, this is because the article did not point out that they were blinded. Again the control groups in the trials frequently gave education making them likely to know.
Although it is not indicated in the article, follow-up might not have been complete. This is due to the withdrawal of some patients from the program. On the other hand, all other patients successfully went through the program.
The intervention must have been consistently delivered because the nurses who performed this study were well trained to follow a specified course of action before conducting it. Again follow-up was done and patients were frequently contacted and the record was taken. Another supportive aspect was that the patients were given a manual that helped in carrying out the program.
The article shows that there is a clear description of the measurement instruments that were used. Applicability and dependability of most determinants were explained and the mode of their management was made clear.
The measurement instruments must have been used validly and reliably although the article does not show much of this. Again there is nowhere in the article starting otherwise. The best information that shows that the instruments were used reliably and validly is on the overall results. The precise estimation of the intervention was again difficult to judge.
The intervention effect in this study looked very minimal. The reason for this is the fact that the nurses conducting the study were well trained before the study. Again the patients were given heart manuals which helped reduce the intervention effect.
These results apply to patient care in the sense that they can help patients with similar circumstances. The subjects randomized were many and therefore the validity and applicability show that they may be well represented in other clinical settings.
All important outcomes must have been considered. This is because this study was well planned before its accomplishment. It is even explained that the nurses were adequately trained before carrying out the study. The patients were again given manuals which showed that the research’s outcomes were considered.
The likely intervention benefits are definitely worth the potential harm and costs.
Although it looked costly due to the number of nurse contact hours needed the intervention could be helpful when a large number of patients are used. The nurse visits and calls also proved that the need for doctors could be minimized.
The study also showed that the intervention in its present structure cannot manipulate risk aspects or post-operative appropriateness. On the other hand, the research demonstrates that it could bring about the use of medical-care resources in the waiting period. Such studies can connect to cardiac curing centers to give structured keep fit guidance to improve fitness. The intervention, therefore, looked like it could minimize general healthcare utilization and thus benefits worth the potential costs and harm.
The results showed improvement in the cholesterol and blood pressure levels even if significant differences were minimal. Again cost minimization was realized in the control cluster. This information is found on page 189 where the authors explain that:
For both groups blood pressure and total cholesterol improved (Blood pressure mm Hg (Control −9.11 (CI −4.89, −13.33); Intervention −13.02 (CI −8.76, −U17.29) both pb0.01); total cholesterol (Control −0.20 (CI −0.03, −0.37) p=0.02, Intervention −0.18 (CI −0.02, −0.34) p=0.03). However, there were no significant differences between the groups. Cost minimization analysis showed that the total costs were less in the intervention group due to fewer admissions (total costs £10,754 (3746) v £13,047 (5835), CI −3743, −843; p=0.002) (Goodman et al, 2007).
On the validity or justifiability of the results, we find that in both clusters the pressure of blood and cholesterol level was enhanced.
The control clusters did not vary significantly for any of the significant baseline medical variables. The mean stay was about five and a half months for the investigational cluster (range 1.5–8.9) then 5.4months for the control cluster (range 1.6–11.5). In analysis, “the national waiting list initiatives led to large numbers of patients (23%) receiving their operations before the three-month measurement point and so the admission point was used for the main analysis” (British Cardiac Society Party, 2002). This is mainly for the reason that, data were available for most patients at this position. For the control group, it was at 98.9% while in the experimental group it was and 95.7%. The purpose of the treatment principle was employed. So generally the treatment effect was large. This was as per the information extracted on pages 194 and 195 respectively.
In table 2 on page 192, the results displayed significant information on improvements from baseline to surgical procedure. On the same page, the authors indicated that “the numbers of smokers were too small for analyses to be meaningful and the change in glucose levels was not significant at any measurement point” (Prochask, 2002). On the quality of life scores, the proof was not found on the effect of the intervention by several measures apart from a statistically borderline development on the bodily quality of life as deliberated by the SF36. This was however not restricted to the cardiac-specific CROQ scores.
This type of program did not impinge on the development of the research’s risk factors throughout the pre-operative wait for the surgical procedure. Nevertheless, the result suggests an effect on medical care utilization with minimized charges for inpatient admittance throughout the waiting period. This is, “in the intervention arm which may be due to the homecare visits. It again requires further exploration in future but patients may have chosen to discuss these problems with the homecare team rather than go to their local hospital” (Linden, 2005). It also duplicates the ‘Heart Manual’s’ assessment which displayed little use in GP and clinical services during home support given by nurses.
Another aspect which may have impinged on the results was the time presented for the intervention. This is because; researchers projected a longer time to make available the intervention prior to surgery.
This might have been the first RCT to put into practice a nurse-coached, home-based IMT intervention. Further research and studies to test efficiency in more patients might be required before making the decision to put into practice this type of intervention in medical practice.
Consequently, in a synopsis this research verified that the intervention in its current structure does not influence risk aspects or post-operative suitability. However, the study shows that it could result on the patients’ use of medical-care resources in the waiting phase. Such a program in future could connect to local cardiac treatment centers to offer structured keep fit training to enhance fitness. Again the nurse-led program was not or did not look like it could minimize risk factors preceding the cardiac surgery (Rickham, 2004). On the other hand, the intervention looked like it could minimize general healthcare utilization.
References
British Cardiac Society Party. (2002). Cardiac rehabilitation services in the UK. The heart journal 73 (4), 1-2.
Cullum, N., & Ciliska, D. (2008). Evidence-Based Nursing. An introduction. Oxford: Blackwell.
Goodman. et al. (2007). A randomised controlled trial to evaluate a nurse-led programme of support and lifestyle management for patients awaiting cardiac surgery ‘Fit for surgery: Fit for life’ study. European Journal of Cardiovascular Nursing 7 (2), 189–195.
Linden, B. (2005). Evaluation of a home based rehabilitation program. London: McGraw-hill.
Prochask, J. (2002). Stages and processes of self change on smoking. J consult cln psycho 51 (6), 152-168.
Rickham, PP. (2004). Human Experimentation. Medical Journal 42 (3), 432-463.