Experimental Evidence and Evaluation - Assignment Example

EXPERIMENTAL EVIDENCE AND EVALUATION Critique The of this research in this application is "Patient preference and clinical outcomes in patientsreceiving physiotherapy for back pain: a randomised controlled trial." This study is a proposal for application of a research grant, and therefore, there is need for critical evaluation of this proposal to examine whether really this study will be feasible. In order to do that, an evaluation of the methodology would be done to see if it has been designed adequately appropriately to answer the research question. Moreover, since it is a clinical experimental trial, whether the end points will be achieved or not is the most important question, so it can have significant input in clinical practice. The title is concise in that it provides information in brief. From the title, there can be idea of the objective, but no idea about the content is available. However, the proposer clearly indicates the research approach used, which is randomised controlled trial. All studies can be categorized as experimental, quasi-experimental/pre-experimental, or non-experimental in design. From the title it appears that the researcher wants to design an experimental study with the purpose of determining the clinical outcomes of physiotherapy and related patient preferences for treatment of back pain. This is a clinical trial, and these studies are designed to determine whether the hypothesised clinical intervention will be effective. In this case, the trial was conceived since in mechanical back pain, surgery is never offered as an option of treatment. This trial was proposed with the purpose of comparing two treatment strategies, namely, physiotherapy for back pain chosen by patients and physiotherapy for back pain offered by the physiotherapist in alleviating back pain of the patients. If one goes back to examine the title again, this comparison through randomised controlled trial is not explicit in the title, so there is an obvious confusion on the aim, purpose, and objective of the study. This study has been proposed to be done in a single physiotherapy department, and therefore this is not a multicentric trial. This reduces the authenticity of the findings in terms of internal and external validity. As a result the reliability will be compromised, and hence evidently there would be lack of generalisability of the outcomes. Thus, applications of the findings in the clinical practice would be severely limited even through the study is able to generate some valid results (Rumoren and Sunday, 2006). Brief Synopsis of the Project Low back pain is a problem with considerable disability with direct cost implications in terms of NHS resource utilisation. Due to its common occurrence, many patients present to GP and utilise other health and social care resources. It is also an important cause of disability, and thus there are implications of work hour losses. From the NHS point of view, low back pain has been demonstrated to represent both public and private costs. Other than the conventional treatment modalities, it is now known that prompt physiotherapy is an acceptable service for these patients. For the physiotherapists, the problem remains which could be the best form of therapy, which the patients choose or what the physiotherapist advises. In many cases, low back pain has no associated organic problem, and thus recovery cannot be ascertained by any investigative objective means. Moreover, pain has a subjective element. Therefore assessment of recovery can be done with a validated instrument, and if this validated assessment instrument is used to compare the responses of two groups of patients with two different interventions, then a trial with randomisation and clinical control would be the best way to determine the intervention of choice in such cases. This part of discussion indicates that the proposer has chosen a clinical problem, the decisive plan of action on which is unknown. The clinical problem here is low back pain and the appropriate physiotherapeutic approach. The researcher does not mention the research question, but it is evident that the question is what physiotherapy approach is best suitable for patients with back pain. As indicated earlier, despite no definitive wording of the hypothesis or the research question, if they are worded properly, this proposed research, however, tends to address the clinical problems of treatment of low back pain in physiotherapy practice. If this grant is allowed and the research is pursued, the practicing physiotherapist would eventually know the better approach out of these two to apply in practice. Given the burden of low back problem on NHS, this can also provide a solution to the NHS issues concerning low back pain. Critique of the Design Purpose To measure the clinical outcome of allowing a patient choice of physiotherapy with low back pain referred to physiotherapy outpatient services. Therefore, there is an intervention, which would be allowing the patients to have physiotherapy regimen of choice in the physiotherapy department with the expected endpoint or outcome being satisfaction in treatment and alleviation of back pain through measurement of improvement in Oswestry Disability Questionnaire (ODQ) scores. However, from the description of the objectives as mentioned, on completion, there will be measurable evidence of the effects of one approach in a specific condition. If the objective for this trial is to provide information and evidence related to patient-chosen physiotherapy on clinical outcomes in management of back pain, then the design is faulty. The objective seems to be accurate due to the fact that care providers will be interested in the results only when they can use the research findings to plan care and plan provision of information to the patient so they can make an informed choice. The design must also incorporate data collection and data analysis in such a form that the mechanism of choice impacting and influencing clinical outcomes. These mechanisms may be useful in designing future studies to probe the problem further so new dimensions of care provisions become explicit. This clinical trial should also demonstrate the use of traditional trial methods in unconventional situations. However, while designing, the researcher thus proposes to make a comparison in the improvements in ODQ scores of back pain patients, who chose physiotherapy on their own to that of patients who were allocated therapies by their physiotherapists. This seems to be incongruent with the purpose and objective of the study in that rather than measuring comparative impact of the two forms of therapies, which are hypothesised to produce outcomes in a nearly similar fashion, allocation to groups and administrations of physiotherapies become the principal outcomes, which can have many confounding factors compromising the internal validity of the data. Moreover, these two parameters may not need a statistical or quantitative comparison. When randomised control trial is the design in the clinical settings, the clinical outcome is the goal out of two different interventions. Moreover, a control group is needed for elimination of bias. Randomisation involves choosing subjects in fashion where there is no pattern. Thus every subject may be assigned to any group. This would ensure absence of systematic bias, and the dependent variable would be unaffected. In research methodology, this is the most dependable method of equalizing two groups. Randomisation entails two separate processes: (a) random selection of subjects from smaller magnitude and (b) random assignment of subjects to treatment and control conditions. In the context of this study, this involves allocation of subjects to intervention and control groups by using random decisions. Randomisation is the strategy that is assumed to allow blinding of the observer so the observer bias does not play as a threat to the study. Moreover, uncontrolled confounding variables are major threats to the internal validity of the study. Randomization reduces this effect. Unless confounding factors are controlled in the design of a study while sampling or while analyzing the data, distortion in the estimate of the effect will result. A confounding factor operates through its association with both the dependent and the independent variable by the threat to produce an indirect statistical association. There is no such mention here, and therefore, this raises a doubt as to whether this design would be sustainable. There is no mention of sample size, process of recruitment, ethical aspects, and methods that make the sample size valid with consideration for attrition. In effect, there is no methodological rigour, and usual classification of intervention and confounding factors is reversed in the design. Thus this methodology would not be suitable for the study and hence is not tenable (Cook et al., 2008). Part B Proposal Title Patient preferred physiotherapy and clinical outcome in physiotherapy for low back pain: A randomised controlled trial Brief Low back pain causes disability and a public health problem recognised in the NHS. For those who work, back pain is the commonest cause for seeking healthcare. This involves costs in terms of healthcare expenditure and loss of working hours. Physiotherapy has been recognised to be an important and effective treatment strategy for low back pain. Physiotherapy in this condition is delivered through a regimen of mobilisation of joints, advice on posture, and exercise programmes. The magnitude of the problem would be understood from the fact that the British NHS treats about 1.3 million people in a year. On the other hand, the role of routine physiotherapy in its treatment has very weak support from evidence. Therefore, as a practice parameter, the physiotherapists must know which therapy is effective in alleviating low back pain. Hypothesis Subjects suffering back pain have preferences for treatment, and pain being a subjective sensation subjective sense of improvement has important bearing on outcome. It is hypothesised that if the patients are allowed to receive physiotherapy of their choice, the therapeutic benefit would be more than that with the therapy provided by physiotherapists, which they may not favour. Therefore patient-preferred physiotherapy regimens have better outcome in treatment of back pain than physiotherapist chosen physiotherapy. Objectives The objectives of the proposed study which attempt to test this research hypothesis are 1 Comparison of changes in ODQ scores between patient-chosen physical therapy to that for patients with physiotherapist-allocated conventional physiotherapy. 2 Identification of patient characteristics influencing the clinical outcomes from physiotherapy preferences. According to the hypothesis, knowledge about locus of control and other important factors may indicate future studies (Velanovich, 2007). Methodological Parameters This is a clinical trial conducted in a physiotherapy department. It is not a multicentric trial. Study Design Back pain is helped best with physically activity of the patients. Physiotherapy is a recommendation. The effectiveness of patient chosen physiotherapy will be compared with that of physiotherapy allocated by physiotherapist. Since the current usual practice is to have physiotherapist allocated physiotherapy for low back pain, the latter group will serve as the control. The protocol for the intervention arm will allow the physiotherapist to select the physiotherapy regimen according to the choice expressed by the patient. The patients will be recruited from the physiotherapy centre in the health service provider where the patients were designed to receive the physiotherapy treatment. The patients will be provided with information regarding available exercises on low back pain. This will allow all the patients to be exposed to different choices and ability to exercise free will regarding participation and choice of exercise. When the patients who are all potential recruits come for physiotherapy, before the first ever session, they will be subjected to a short interview, where their preferences of physiotherapy will be asked. These preferences will be intimated to the treating therapist with instructions to pursue with the patient's choice unless there is a chance of harm with that particular exercise. This particular step decides the intervention group, and since the patient's choices are not known early on, this also serves as an acceptable method of randomisation (Coughlan et al., 2007). The patients who would not choose any particular physiotherapy regimen would serve as the control group since this is the usual form of treating such patients when the physiotherapist allocate the treatment to such patients as deemed suitable. This means, according to the normal practice, the control group will be provided information and treatment with physiotherapy sessions offered and delivered by the physiotherapist. It is to be mentioned that both groups will access only available therapy within the department, ensuring randomisation in this design with no scope for bias that may affect results (Sousa et al., 2007). The comparison thus can be quantified between patients who can choose an available treatment and those treated conventionally at the discretion of the physiotherapists, assuming both these formats would yield effective management of back pain. The allocation plan and modes of interventions would lead to the main differences in the measured outcomes. These confounding factors would be the main effects to be quantified in this trial. Participants and Eligibility This study has been designed as a unicentre investigator-blinded clinical trial with randomisation and control. All the patients will be approached with both verbal and written explanation of the research and the outcome it is going to generate. Since all the patients are presenting for receiving some kind of physiotherapy, apart from assumptions that both intervention and control group treatment would result in benefits for their back pain, there is no ethical twist in this study. Participants will be selected from the patients attending the NHS outpatient physiotherapy clinic for treatment of back pain. The informed consent may lead to a bias in the patients leading to alteration in the data collected due to the subjective nature of the outcome. This may alter the results of the study. The control participants may exercise more choice. Many participants with strong preferences may decline participation, leading to compromise in the effect size, overshadowing the true effect of the interventions. A modified form of consent has been proposed. The patients will be invited to participate in the study by an explanatory letter. This will be posted to them. This will accompany the initial questionnaires for assessment. The patients who would return the completed questionnaires would obviously participate in the study, and this will be regarded as consent. These consenting subjects may then be randomly allocated to the groups with extra information and control groups. These subjects will be blinded of the procedure of randomisation and would also be blinded of the groups of comparison again to avoid bias. All subjects will then be asked of informed consent to be recruited in the study before they are included (Driessnack et al., 2007). This will mean the intervention arm of the participants may choose conventional treatment as an option. This means they will be recruited in the control arm of the study indicating their dissent to the intervention. Harm risks from the ethical point of view due to information sharing and requesting preference would be practically nil. There would be provisions for no new or different therapy. At any point in time, the patients will have liberty to opt out of the study, and it will be revealed clearly to them that their decision to not to participate will not in any way affect their care. At any cost confidentiality will not be violated, and there would be no coercion to participate in the study (Richardson and McMullan, 2007). Inclusion and Exclusion Criteria The inclusion criteria will be any patient over age 18-65 years will mechanical back pain for a duration of six weeks at the minimum with or without signs of neural compression indicated by leg pain or other neurological signs without gender or ethnic considerations. These patients will be recruited from the referral lists of patients referred by GPs or consultants with mechanical low back pain. A group of research therapist will examine the patients to check eligibility to participate in the study. There will be screening questions to exclude serious disease, where physiotherapy may be risky. These patients will be excluded from the study. These conditions will include symptoms suggestive of systemic diseases, rheumatic diseases, ankylosing spondylitis, spinal tumors, spinal infections, prior spinal surgery and rigidity, gynaecological disorders, and physical problems or their treatments in the past month. The patients with the need for functional restoration will also be excluded. Moreover patients referred but unable to attend outpatient treatment will be automatically excluded from this study. The patients who are not able to understand written English would not be able to understand the information and the questionnaire, so will not be able to respond, and these patients will also be excluded. Those who have received physiotherapy in the last 3 months, recent back surgeries, diagnosis of other causes of back pain as mentioned above, and a running litigation over back injury will be excluded due to difference of effect. This may lead to an imprecise estimate of effect and would constitute small numbers, leading to inadequate sample size for subgroups (Fairbank et al., 2005). Sample Size The first step towards this is an adequate sample size which will be assigned as mentioned below. A. Intervention arm : 125 B. Control arm: 125 Sample Size Calculation This sample size was determined based on the minimum effect size to validate the hypothesis. The ODQ scale will be utilised to measure and compare the outcome. The minimum effect size will be indicated by a 4 point difference in the ODQ scale, which will be a clinically significant difference. The standard deviation for the within person difference in ODQ scores is expected to be 8 points. The number will be given by the formula N = {4* [(z = 0.05) + (z = 0.2)]} / [(4/8)^2] = 127 There will be a 10% crossover and 80% followup doubles which calculate to a total of 250 subjects, which will be divided equally between the two arms of the study. As evident from the formula, an error of 5% and a error of 20% have been taken into consideration as level of significance in sample size calculation. Without statistical methodology, summarization of the data, determination of sampling error, and analysis of the results to determine statistical significance are not possible. This does not mean that in designing the methodology, only statistical methods are of importance (Redman, 2006). To test the hypothesis adequately, power analysis is used to estimate sample size needs. Although any study would include the largest possible sample size to effect generalizability, every time the researcher calculate a percentage or an average based on sample data, they are estimating a population value. For this reason power analysis is used that considers effect size that expresses relationship among research variables. Was there formal statistical input to the overall study design In this study design, there was formal statistical inputs from the qualified authority, the details of which have been given below. If yes, please specify the name of the advisor: ----- The advising department or institution: ----- Main Outcome Measures In the absence of reliable laboratory tests or other imaging studies for assessing response to therapy for back pain, an instrument, the Oswestry Back Pain Disability Questionnaire (ODQ) will be used as the main outcome measure. Another instrument, the Quebec back pain disability questionnaire can be used. It is acceptable since it demonstrates consideration of psychometric properties. Both these instruments may be used and the results compared (Poolman et al., 2007). The secondary outcomes will be measured by patient-rated satisfaction scored with the care they received. At the end of the therapy, both the patients and the therapists will rate their perceived degrees of improvements. A visual analogue scale or Likert scale will be used to rate strength of belief with 0 indicating minimum belief and 10 indicating maximum belief in different forms of therapies (Tiffany, 2006). The Internal Control Index questionnaire will measure the health locus of control (Velanovich, 2007). This is a personality trait which may affect the perceived results of the intervention, thus quantifying its effects. A self-completed questionnaire will be sent and returned by post, and the above results will be recorded. All the patients will be followed up and reminded by post and telephone or personal visit up to four times or until they express their desire to withdraw from the study. Reference List Cook, DA., Bordage, G., and Schmidt, HG., (2008). Description, justification and clarification: a framework for classifying the purposes of research in medical education. Medical Education; 42(2): 128-133. Coughlan, 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. Driessnack, M., Sousa, VD., and Mendes, IA., (2007). An Overview of Research Designs Relevant to Nursing: part 2: qualitative research designs. Rev Lat Am Enfermagem; 15(4): 684-688. Fairbank, J,, Frost, H., Wilson-MacDonald, J., Yu, L-M., Barker, K., and Collins, R. for the Spine Stabilisation Trial Group, (2005). Randomised controlled trial to compare surgical stabilisation of the lumbar spine with an intensive rehabilitation programme for patients with chronic low back pain: the MRC spine stabilisation trial. British Medical Journal; 330: 1233. Poolman, RW., Struijs, PA.A., Krips, R., Sierevelt, IN., Marti, RK., Farrokhyar, F., and Bhandari, M., (2007). Reporting of Outcomes in Orthopaedic Randomized Trials: Does Blinding of Outcome Assessors Matter Journal of Bone and Joint Surgery. America; 89: 550 - 558. Redman, BK., (2006). Review of measurement instruments in clinical and research ethics, 1999-2003. Journal of Medical Ethics; 32: 153 - 156. Richardson, S and McMullan, M., (2007). Research Ethics in the UK: What Can Sociology Learn from Health Sociology; 41: 1115 - 1132. Rumoren, M. and Sunday, J., (2006). Where is the humility for the limitations of research Clinical Infectious Diseases; 42(6): 888-889. Sousa, VD., Driessnack, M., and Mendes, IA., (2007). An overview of research designs relevant to nursing: Part 1: Quantitative research designs. Rev Lat Am Enfermagem; 15(3): 502-507. Tiffany, JS., (2006). Respondent-driven sampling in participatory research contexts: participant-driven recruitment. Journal of Urban Health; 83(6 Suppl): i113-i124. Velanovich, V., (2007). Behavior and Analysis of 36-Item Short-Form Health Survey Data for Surgical Quality-of-Life Research. Arch Surg; 142: 473 - 478. Read More