Review of a piece of research based evidence relevant to health care practice

Reviewing research based evidence is essential to any health care practice in providing the best possible care. Evidence based practice (EBP) ensures that the healthcare given is effective, up-to-date and grounded in evidence appropriate to the situation. There are many different sources that can be searched for research based evidence, the most useful being the Cochrane Collaboration and the NHS Centre for Reviews and Dissemination (NHS CRD) at the University of York, evidence based journals and computerised databases (Thompson, 1999).

The topic of study under review is effectiveness and skin tolerance to hand decontamination procedures. In order to find published research it is first important to generate a clinical question so that the key components of the question guide the search strategy. A clinical question was created by defining the problem and intervention, giving a comparison and by hypothesising a clinical outcome. From this question keywords were selected for use in implementing a broad search of the available data. This search was then narrowed through a filtering process where evidence was rejected if inappropriate or of least relevance (Flemming, 1999). Limits used in the filtering process were the type of study (only Randomised Controlled Trials included), when they were published (since 2000) and the language they were published in (English). In order to find published data on the topic area of hand decontamination several online databases were searched (EBSCO Host EJS, CINAHL and Medline) using the key words of 'hand decontamination', 'hand washing' and 'skin tolerance'. This search provided the article under review which is a paper written by Winnefeld et al (2000) 'Skin tolerance and effectiveness of two hand decontamination procedures in everyday hospital use', published in the British Journal of Dermatology.

The study by Winnefeld et al (2000) is concerned with nosocomial infections caused by micro-organisms disseminated via the hands of nursing staff within a hospital environment. Additionally, the paper questions whether the damaging effects of hand washing on skin may adversely influence hygiene practices of health care workers. As damaged skin may also become more easily colonized by micro-organisms, the paper hypothesises that a hand disinfectant needs not only to have an anti-microbial effect but also an optimum degree of skin tolerance is required to obtain efficacy. A key question is whether a frequent application of a product as dictated by hygiene requirements can be tolerated from a dermatological point of view. This is examined by comparing two widely accepted hand hygiene measures, namely a non-antiseptic liquid soap (Savodoux) and an alcohol based antiseptic hand rinse (Sterilium), on their anti-microbial efficacy and their effect on skin tolerance. The Study is quantitative, randomised controlled trial (RCT) using a pre and post test design.

The method of sampling is incidental or opportunistic which means that participants were a selection of the most easily accessible members of the target population (Polgar and Thomas, 2000). Indeed, 52 nurses out of 78 volunteers were selected for the study, with 51 completing the study. It is difficult to asses the appropriateness of a sample size without knowing the demographics of the target population, as the optimum sample size is one which is adequate for making inferences from the sample to the population (Polgar and Thomas, 2000). Furthermore, without any confidence intervals reported it is hard to determine, retrospectively, the adequacy of the sample (Thompson, 1999). Indeed, sample size is significant as too small a sample may lead to sampling error which is reflected in the discrepancy between the true population parameter and the sample statistic (Polgar and Thomas, 2000). As the size of the sampling error generally decreases as the size of the sample increases (Thompson, 1999c) there should be substantiation for the sample size, yet this was not provided by Winnefeld et al.

In this paper, it is not clear who the target population are, yet it appears to be all health care professionals with patient contact within a hospital setting. However, the sample only consists of nurses and nursing assistants and therefore there is a huge over-representation of nurses and a huge under-representation of other health care workers. Furthermore, the age and sex of the participants are said to be evenly distributed, yet the age range of the participants was not given and the male/female ratio appears skewed with 49 females and 2 male subjects participating. Therefore, the sample may have been more representative if the method of quota sampling had been adopted, where age, sex and profession could proportionally be represented within the sample.

When using an experimental design, care must be taken in the assignment of participants to avoid significant initial differences between intervention groups (Polgar and Thomas 2000). Randomisation of participants is meant to eliminate any bias that may occur in this process, and the study by Winnefeld et al (2000) did implement randomisation. However, it is not clear how the groups were randomised and subsequently this can hinder the understanding, evaluation and perhaps replication of this research project. To eliminate any bias it is further recommended that whenever possible studies should be double or single blind (Polgar and Thomas, 2000). This is when the participants, experimenters or observers are not aware of the aims and predicted outcomes of the investigation. This is to safeguard the internal validity of the experiment by defending against observer bias and expectancy effects (Polgar and Thomas, 2000). Blinding was not employed in the study by Winnefeld et al (2000). It is clear from the very nature of the intervention that participants could not be blinded as to their intervention group. However, it could have been possible for the experimenter or observers to be blinded to the randomisation of participants, though this was not stated to have happened. As such, the internal validity may be threatened.

General

This introduction does several different things well. It provides a rationale for the piece of work by discussing the general importance of evidence review. It describes the search methodology used by the writer in sufficient detail to allow the method to be reproduced. It provides a concise précis of the research paper under discussion, using appropriate language.

View linked text

Quality: Structure

The first few sentences provide a succinct general introduction to the process of reviewing evidence-based research. Note that in this case the topic of the actual study is not introduced until the second paragraph. The writer is foregrounding the importance of evidence-based practice in healthcare in general before considering a particular case.

View linked text

Function: Describe

This sentence introduces the topic.

View linked text

Before describing the search method, the writer makes a general observation on the importance of generating a clinical question to guide the search.

View linked text

This section gives a clear, step-by-step description of the search strategy that was employed to search for evidence on this topic. As with a scientific method, enough detail has been included to allow a reader to repeat the method, using the same databases, search terms and filters.

View linked text

This section clearly summarises the issues under investigation in the original research paper. Notice that although the summary is presented in the writer"s own words, the language is very close to that of the original research paper and assumes a similarly informed reader. It is good practice to make use of technical terms such as "nosocomial" and professionally-recognised phrases such as "optimum degree of skin tolerance" to show that the writer is comfortable with these uses of language.

View linked text

Details of the research methods are given, again based on the original paper.

View linked text

An accurate description of the sampling method has been provided along with a definition of this method in the writer"s own words, but supported with an appropriate reference to the general literature on research methods.

View linked text

The writer moves on to describe the process of assigning participants to groups in a trial, again with reference to the general literature. Although the start of a new paragraph can indicate a new topic, this could have been made clearer using a short introductory sentence such as: "There may also be weaknesses in this study concerning randomisation."

View linked text

The blinding process is described with reference to the general research literature.

View linked text

Function: Analyse

The appropriateness of the sample size in this specific case is considered with reference to the general rule established. An appropriate reference has been used to support the point that is being made.

View linked text

Additional analysis of the study clarifies that there are weaknesses in the sampling strategy and that this makes it difficult to interpret the findings accurately. References to the general research literature substantiate the points being made.

View linked text

The writer makes more specific criticisms of the sampling strategy undertaken by the original study. Some assumptions are being made here, such as that other health care workers should have been included, and these could perhaps have been made explicit. It is interesting that the language used by the writer becomes less measured ("huge") when assumptions are being made that are not strictly based on the evidence of the research paper.

View linked text

It is clear from the analysis exactly how the original study falls short: not in failing to randomise but in failing to clarify how this was done. Consequences of this failure are spelled out. The word "subsequently" is not necessary in this sentence and sounds as though it has been used as a more impressive synonym for "so" or "therefore", which would have been better.

View linked text

An analysis of the research study finds evidence of a weakness and suggests how this could have been overcome.

View linked text

Function: Evaluate

This is an excellent closing sentence, summarising very succinctly the negative consequences of the weaknesses that have been analysed. "Because of this" would be a slightly more correct joining phrase than "As such".

View linked text

This concluding judgement on the weaknesses of the sampling strategy usefully focuses on how the original researchers could have improved their study.

View linked text