Using evidence-based information is a key nursing practice. An important instrument to building up evidence-based practice is a critical review of current research. This essay aims at critically reviewing the article (Randomized double-blind placebo-controlled donepezil augmentation in antidepressant-treated elderly patients with depression and cognitive impairment: a pilot study) by Pelton et al (2008) using Parahoo critical framework.
Using the current best evidence is a key practice of nursing. Therefore, nurses must-read research critically, which is a difficult process. Despite being deterring yet, it is an essential skill to develop nursing ability to identify the best practice (Coughian et al, 2007).
Providing best practice does not necessarily mean implementation in daily practice, however, transference of best practice remains the first step in the process of application evidence of best practice inpatient care. This is the gap between evidenced best practice and reality practice, nursing attitudes and perceptions of utilizing research s major contributing factors to the gap. Evidence best practice can be developed from systematic reviews, with randomized controlled trials being among the most valuable research to study (Rycroft-Malone et al, 2004). A structured framework for an article critique as suggested by Parahoo and others (after Ingham-Broomfield, 2008) is a tool that incites questions leading to better assessment of the value of a research paper. The primary aim of this framework is to question each section of the paper to achieve better interpretation of the content (Ingham-Broomfield, 2008). The aim of this essay is to review critically the article of Pelton and others (2008) about administering donepezil to improve cognitive functions in the depressed elderly, using Parahoo structural framework as described by Ingham-Broomfield (2008).
Research articles’ titles can be descriptively stating the focus of research or conclusive stating the authors’ conclusion. In either case, the title should be clear conveying the article’s hypothesis and methods. It should also contain enough keywords to facilitate search in databases (American Psychological Association, APA, 2001, pp. 3-30). The title of the article at hand fulfills these requirements.
Authors and date
The author’s institutional affiliations are relevant to the subject of the paper and there is no delay between accepting the paper and publication pointing to its relevance to current practice (Dawson, 2002).
Although the abstract is well structured stating the objective of the work, methods used, results and conclusions, however it was not preceded by one or two sentences as introductory statements summarising current and, or unsettled knowledge (Alexandrov and Hennerici, 2007).
Introduction and statement of the research problem
In a research article, the introduction is the part that provides an outline of the subject studied highlighting the theoretical framework and the need to study this particular subject. Thus, introduction of a research article has three main components a brief description of the topic, statement of research problem and the background literature (Fawcett and Garity 2009, p.37).
The research problem is the part that attracts the reader’s attention to the scope, impact and the status of the problem studied (Fawcett and Garity 2009, p. 7).
Pelton et al (2008, p. 670), clarified, in the introduction, that depression and cognitive impairment are the most common neuropsychiatric disorders in the elderly besides, being a common comorbidity. Further, they clarified that building a long-term treatment strategy for both conditions should have a better impact on disability and overall health mental outcome. They also displayed evidence that acetylcholine esterase inhibitors (including donepezil) have the potential to treat mild to moderate cases of cognitive impairment. The authors limited their research problem to examine the results of a previous study that reported no or limited improvement using donepezil for treating cognitive impairment in the elderly. For this purpose, they conducted a randomized double-blind-placebo controlled study to assess the benefits of donepezil for the same period (12 weeks), with a sub-sample of continued treatment for 8 months.
Review of the literature mainly aims at defining or building up the research questions, recognizing and presenting methods of data collection. Secondly and of equal importance, it aims at defining research gaps linked to the study and how to compensate for them. Therefore, literature review should display appropriate depth and width of information about the subject of the study (Coughian and others, 2007). According to the WHO report (2000, p. 11), a literature review in a clinical trial should also identify the current evidence of efficacy and safety of the suggested intervention.
Analyzing the literature review of the article at hand, all sources are peer-reviewed journals, and from primary source data; however, many of the references date way beyond five years of the publication date. Some authorities consider this a defect especially in the absence of a historical introduction (Coughian and others, 2007).
In this context, the authors did not clarify the theoretical framework behind choosing a randomized double-blind-placebo controlled study or how did they overcome the ethical debate on placebo trials (Regan et al, 2005). The authors also did not overview the theoretical framework of using acetylcholine esterase inhibitors in treating cognitive impairment. The authors did not comment on sertraline-donepezil drug-drug interaction or augmentation of side effects. It is true that both drugs have different mechanisms of action (sertraline being a serotonin reuptake inhibitor) and no reports of drug interaction exist. However, both drugs are metabolized in the liver, and have common side effects like headache, anorexia, nausea, diarrhea and insomnia (Spina and de Leon, 2007).
In 1997, Parahoo (p. 142) described research design as the plan to answer the questions of how, when and where data are collected and analyzed. Thus, research design enables the reader to assess the work done and to repeat the study if they need to. Therefore, authors of research articles should describe research design; associated limitations and describe the research participants. Description of participants should include total number, number of participants in assigned groups, age and gender distribution, method of selection. It should also include sample characteristics of importance to the study like demographic characteristics (Hittleman and Simon, 2006).
Pelton et al (2008, pp. 671-672) described criteria to include or exclude patients from the study. They stated that all 23 elderly patients included in the study had different phases (A, B, and C). Age, gender and other characteristics of the sample studied were in mean and standard deviation but not as range in table 1 (p. 672). This indicates the inability to infer results as they link to age or gender. The authors explained the donepezil treatment plan as regards doses and treatment duration for the three phases of the study. Details of the number of patients assigned for each treatment phase whether for placebo or actual treatment. Figure 1 outlines the number of patients assigned to each treatment phase sample whether for placebo or actual treatment. However, the authors neither mentioned nor referred to the details of their sample in the text nor specified the number of patients and criteria of selecting placebo group; they also did not clarify whether placebo patients gave their consent on the trial.
A sample is a part of a subset of the population that is the total number of population units from which data can be collected (Parahoo, 2006 after Fawcett and Garity, 2009, p. 135). The authors did not define the method of randomization in the sample selected nor did they define detailed information about the sample. On the other hand, they defined but did not describe the population from which the sample is selected. Based on the above discussion, an example of identifying the sample in the research article at hand would be (Fawcett and Garity, 2009):
Sample: Composed of male and female elderly (50 years or more) depressed patients. Sample size is 23 patients. Inclusion criteria: Diagnosis of major depression, dysthymic depression or depression not otherwise specified (according to DSM-IV) patients included an HRSD (Hamilton Rating Score for Depression) equals or more than 14. For cognitive impairment, subjective cognitive complaints and impaired performance were confirmed by a brief battery of tests, with all patients showing no activity of daily living deficits. Exclusion criteria: These were diagnosis of dementia (or other DSM-IV axis 1 psychiatric disorder), or patients with medical or neurological causes for mood or memory impairment. Patients on antidepressant treatment before the trial stopped their medication 1 week before joining in.
There is no predetermined formula for proper sample size needed for research, however the larger the sample the lesser chance to type I or type II errors on testing the null hypothesis (Salagnik, 2006). Since the authors selected their study to be a pilot study, there was no need for a larger sample.
In summary, since the population and sample characteristics are unclear, the methods of randomization and sample selection are indistinct and the sample size is small, this may cast shadows on operational adequacy of the article, or simply how good sample information is (Fawcett and Garity, 2009).
Instrument, measures and variables
The methods section of a research article should include adequate description of the research instrument (s) used, how findings are measured and the variables measured. These relate to the validity and reliability of results (Hittleman and Simon, 2006). A research instrument is the study tool used for assessment of variables or factors that need measuring; they can be self-administered questionnaires or a battery of psychological tests. Measurement is the assignment of numbers to represent the amount of attributes present (Beck and Dennis, 2003, p. 413). In quantitative research, a variable is a feature measured (like weight, depression, cognition…) (Beck and Dennis, 2003, p. 29).
Selecting proper research instrument will produce a reliable measurement of variables and reflects study validity and reliability (Beck and Dennis, 2003, p. 416). Reliability of a research instrument expresses the accuracy of measuring a target variable. Thus, the less variation the instrument produces in repeated measurements, the higher the reliability and the more consistent the results are. Reliability is also a measure of the accuracy of results (Beck and Dennis, 2003, p. 416). Validity is the degree of the instrument to measure what it is designed to measure that is the scores obtained truly reflect the extent of change of the variable measured (like depression). Validity and reliability are related qualities of a research instrument since an instrument cannot be valid but unreliable, and an unreliable instrument produces too many errors to consider it valid (Smith, 2005).
Pelton et al (2008) used two research instruments, the 24-item Hamilton Rating Scale for Depression and the neuropsychological testing (NPT) battery. Ryder et al (2005) have examined reliability, validity and item characteristics of the 24-item Hamilton Rating Scale for Depression and inferred that the test is suitable for assessment of depression and statistical analysis. Neuropsychological testing battery (NPT) or psychometric testing is also a reliable and a valid method to assess cognitive-behavioral functions and abilities. It is an objective method to evaluate cognitive but not mental health functions (Goldstein and Silverman, 2005).
Results section should provide a complete account of, and limited to all relevant findings obtained from the study (APA, 2001).
The authors used ANOVA (analysis of variance) and ANCOVA (analysis of covariance) statistical tests to show significant effects of using donepezil to treat cognitive impairment. Both tests are forms of linear models with a continuous outcome variable (improved cognition) and continuous categorical input variables (sertraline, donepezil, and placebo) in ANOVA and some categorical input variables (a combination of any two). However, the hypothesis of ANOVA or ANCOVA is the equality of mean effects of input variables. This cannot be the case in this paper as sertraline, donepezil, and placebo have different targets (outcome variables), different clinical effects and different mechanisms of action (Hardle and Hlavka, 2007, p. 39).
In addition, despite the abundance of independent variables (age, gender, drug dose, duration of cognitive impairment) affecting the dependent outcome variable (response score), yet the authors did not perform correlation or regression statistics.
Correlation statistics measure the strength of association between two variables that is a measure of the certainty of prediction of one factor (like drug dose) from another (response score). Regression, on the other hand, uses one or more independent variables to predict the value of a dependent variable (Hardle and Hlavka, 2007, pp. 35-39). In the light of the previous discussion, the study shows improve cognitive function with the use of donepezil, but did not provide any relationship between improvement and any of the other variables.
The discussion section represents the researchers’ forum to discuss the contribution of others and to suggest reformulation of the current theory (APA, 2001). In this context, Pelton et al (2008) described the contributions of other investigators, outlined the limitations of their research, and considered this study an introductory for further research. The authors, because of the small sample size did not provide a conclusion to their study.
Based on Parahoo’s framework for a research article analysis, Pelton et al (2008) studied the effect of donepezil on improving cognitive function in the depressed elderly. They selected a randomized double-blind placebo-controlled design of a clinical trial. The authors did not clarify the theoretical framework of their study, nor did they define the principles of randomization in their small-sized sample. They selected 24-item Hamilton Rating Scale for Depression and neuropsychological testing (NPT) battery as research instruments, both instruments proved valid and reliable. For data analysis the authors relied on ANOVA test, which is partially suitable for their data and did not provide further correlation or regression analysis of their data. The result of this study suggests donepezil can be a successful adjuvant treatment in the depressed elderly with mild to moderate cognitive impairment; however further detailed, large-sample studies are needed.
- Alexandrov, AV, and Hennerici, MG, 2007. Writing good abstracts. Cerebrovasc Dis, 23, 256-259.
- American Psychological Association, 2001. Publication manual of the American Psychological Association (fourth edition). Washington, DC: American Psychological Association (APA).
- Beck, CT, and Dennis, FP, 2003. Nursing Research: Principles and Methods. Hagerstown: Lippincott Williams & Wilkins.
- Goldstein, MA and Silverman, ME, 2005. Neuropsychiatric Assessment. Psychiatr Clin N Am, 28, 507-547.
- Coughian, M, Cronin, P and Ryan, F, 2007. Step’ by-step guide to critiquing research. Part 1: quantitative research. British Journal of Nursing, 16(11), 658-663.
- Dawson, C, 2002. Practical Research Methods: A user-friendly guide to mastering research. Oxford, UK: howtobooks.
- Fawcett, J, and Garity, J, 2009. Evaluating Research for Evidence-Based Nursing Practice. Philadelphia, PA: F. A. Davis Company.
- Hardle, W, and Hlavka, 2007. Multivariate Statistics: Exercises and Solutions. New York: Springer.
- Hittleman, D, and Simon, AJ, 2006. Interpreting educational research: an introduction for consumers of research (4 Th edition). Upper Saddle River, NJ: Merrill
- Parahoo, K, 1997. Nursing Research: Principles, Process and Issues. London: Macmillan.
- Pelton, GH, Harper, OL, Taber, MH, et al, 2008. Randomized double-blind placebo-controlled donepezil augmentation in antidepressant-treated elderly patients with depression and cognitive impairment: a pilot study. Int J Geriatr Psychiatry, 23, 670-676.
- Regan, J, Hamer, G, and Wright, A, 2005. The Placebo: Promise and Compromise. Tennessee Medicine, 138-139.
- Rycroft-Malone, J, Harvey, G, Seers, K, Kitson, A, et al, 2004. An exploration of the factors that influence the implementation of evidence into practice. Journal of Clinical Nursing, 13(8), 913-924.
- Ryder, AG, Goldbloom, DS, Schuller, DR, and Bagby, RM, 2005. Use of Psychometric Principles in Evaluating the Hamilton Depression Rating Scale. Directions in Psychiatry, (25), 241-255.
- Salganik, MJ, 2006. Variance Estimation, Design Effects, and Sample Size Calculations for Respondent-Driven Sampling. Journal of Urban Health, 83(7), i98-i112
- Smith, GT, 2005. On Construct Validity: Issues of Method and Measurement. Psychological assessment, 17 (4), 396-408.
- Spina, E, and de Leon, J, 2007. Metabolic Drug Interactions with Newer Antipsychotics: A Comparative Review. Basic & Clinical Pharmacology & Toxicology, 100, 4–22
- World Health Organization (WHO), 2000. General Guidelines for Methodologies on Research and Evaluation of Traditional Medicine. WHO/EDM/TRM/2000.1. Geneva: WHO.