Experimental design evaluation - Essay Example

EXPERIMENTAL DESIGN EVALUATION After an analysis of the grant application for the proposed project “Patient preference ad clinical outcomes in patients receiving physiotherapy for back pain: A randomised controlled trial”, I must regretfully recommend that this grant not receive funding from our organisation at this time. There are several major flaws in both the methodology and the ethical considerations of the proposed study, as well as errors in some of the basic assumptions upon which the proposed methodology described in this grant has been based.

Overall, it does not seem that the methodology suggested will answer the hypothesis for the project, nor have the ethical requirements for the informed consent and protection of the participants been fully met. One major ethical consideration in this proposal is the interview process, which then requests that the treating physician provide the requested treatment to the study group. While the methodology does state that the treatment requested cannot be given if the treating therapist believes that it may cause harm to the patient, it does not make allowances for the patient's best interest as viewed by the practitioner.

Simply because the requested treatment does not cause harm does not mean that it is the most effective treatment for a given patient. One basic assumption for this study is that patient-requested treatments will, in the end, approximate standard clinical treatment practices. However, this assumption puts a great amount of trust and faith in the educational level of the average physiotherapy patient. Physiotherapists undergo years of training, followed by continuing education for the rest of their career.

Patients cannot simply select a treatment, based on at best a limited understanding of the issues at hand and at worst direct-to-consumer medical advertising or anecdotal evidence, and have a practitioner simply provide such therapy. The ethical implications in this situation are quite clear: while the patient's desires should be foremost in the mind of any medical practitioner, it is not ethical to allow the patient to make decisions without assistance in evaluating the available options (Slowther, 2010).

An additional ethical consideration falls under the modified informed consent plan listed in this grant. According to the grant, those consenting patients who are assigned to the intervention group “may choose standard consultation as a real option”. This statement seems nonsensical given the proposed methodology. Once the interview has been conducted and the therapist provided with the results of the interview, the information cannot be rescinded. It would become unethical for the practitioner to ignore the preferences stated by the patient during the interview.

While the patients can withdraw consent for further participation in the study and for the use of their data, the effect of the study would already have occurred and could not be removed. Their care would be altered regardless of their final consent. Another major assumption of this grant is that a “normal consultation”, as it is described in this study, does not allow for patient choice. The converse to the ethical guidelines of best interest described above is the responsibility of a medical practitioner to respect the treatment wishes of a patient as long as that patient is capable of making autonomous decisions (Slowther, 2010; Appelbaum & Grisso, 1988).

Any ethical consultation thus must include a discussion of the patient's choice in their treatment. This represents a major flaw in the methodology of the study, as the only true distinction between the study group and the control then becomes that the patient is directly asked to provide an opinion during the pre-consult interview. The possibility exists that a patient-directed practice will already include such direct questions as a matter of course! This issue would potentially invalidate the results of the study, since in this case there would be no actual “intervention” being performed on the intervention group.

Unless the consultation is being observed by the researchers, which carries with it its own ethical implications for patient privacy, it would become impossible to separate the situations of patient choice from the consultations where the patient simply accepts the base recommendation of the therapist. The therapist being provided with the results of the interviews also stops the study from being blinded. The practitioner is then aware of which participants are in the intervention group versus those patients in the control group.

This knowledge could consciously or unconsciously change the behaviour of the therapist with regard to the patients in the study, the patients in the control group, or the patients in the intervention group. This could also potentially invalidate any results from this study. I hope that the committee requests that the researchers involved in this grant to submit a modified proposal for review. The hypothesis proposed in this grant has great implications for the ethics of patient selection in their own medical treatment, should it be proven that educated and competent patient choice alone can improve patient outcomes for the same therapies being provided.

This study is exactly the type of research that this committee is designed to support. With a modified methodology and the ethical considerations addressed, I would definitely be able to reconsider my recommendation. References Slowther, A. (2010) Ethical Considerations: Respect for Autonomy. Available: www.ethics-network.org.uk/ethical-issues/conscent/ethical-considerations/ (Accessed: February 28, 2011). Appelbaum, P. S. & Grisso, T. (1988) Assessing Patients' Capacities to Consent to Treatment.

IN New England Journal of Medicine, 319. MODIFIED PROPOSAL APPLICATION We appreciate your consideration in requesting an updated proposal from our team. We therefore submit this modified application for your review. Other minor changes have also been made to correct stylistic or typographical errors. Changes are notated with a text style and colour change. APPLICATION FOR RESEARCH & DEVELOPMENT FUNDING SECTION A: SUMMARY Full title of project: Patient preference and clinical outcomes in patients receiving physiotherapy for back pain: A randomised controlled trial Brief (250 word) synopsis of project (give a short summary of the purpose of the research, the potential application of its results and its value to the NHS: The purpose of this research is to measure the clinical impact of encouraging patient choice in the setting of a physiotherapy department.

The randomised controlled trial method will be used to compare the improvement in Oswestry Disability Questionnaire (ODQ) scores for patients participating in choosing which physical therapy to receive for back pain with the improvement in ODQ scores for patients allocated their treatment by the physiotherapist without intervention from this study. The design is a little unusual in that the impact of the therapies themselves is of secondary interest and assumed to be comparable, while the processes of allocation and administration of therapies are the main effects to be measured.

In effect this trial reverses the usual classification of ‘intervention’ and ‘confounding factors’. On completion this trial will provide useful contributions in three areas: 1. It will provide precise evidence of the impact or lack of impact of one simple approach in one condition. It will also provide more general information about the likely impact of patient choice on clinical outcomes. This will help care providers to plan how they provide information and care. 2. It will provide some understanding of the mechanisms by which choice may impact on clinical outcomes.

This may give pointers to future health services research. 3. It will provide an example of, and some lessons to be learned from, a novel use of traditional trial methods. Main applicant: ------ Parent organisation/trust: ----- Name of hospital (if appropriate): ----- Name of university department (if appropriate): ----- Proposed starting date: 1 Jan 2006 Proposed duration (months): 18 Locations where research will take place: Xxxxx Orthopaedic Centre General Practice physiotherapy clinics Summary of support requested at section E: Capital Year 1 Year 2 Year 3 Total Costs ? ? ? ? ?

nil 17376 8176 nil 25552 Other funding received for this project: Grant awarding Body: ------- Year 1 Year 2 Year 3 Total ? ? ? ? 2250 1125 nil 3375 SECTION B: STUDY OUTLINE Research objectives: It is hypothesised that patients have preferences for choice in their medical treatments and that they will gain more benefit when they feel they have had an active role in the selection of their treatment according to their preferences. In order to test this hypothesis the trial objectives are: 1. To compare the improvement in disability scores for patients participating in actively choosing which therapy to use against those who receive an unmodified consultation and therapy plan 2.

To identify personal characteristics that may modify the clinical effect of exercising preference. It is hypothesised that locus of control will be an important factor but other characteristics will be examined to give direction to future studies. Is this a clinical trial? Yes Is this part of a multi-centre trial? No Please give a concise description of the methodological design(s) to be used in your study: Randomised Controlled Trial: with modified consent The intervention group will receive a short informational pamphlet prior to their first consultation appointment.

This pamphlet will describe the therapies offered by the practice, their benefits and disadvantages in treating various types of conditions, and finally include a statement pointing out that patients are always free to request treatment that they prefer several times throughout the text. This method would most likely cause a patient to be more assertive than normal in requesting a preferred treatment, and also cause them to be informed about whatever treatment the therapist recommends for an added feeling of control if the patient chooses to simply follow the therapist's lead.

The additional information provided to the patient does not change the way the therapist behaves, ensuring the patient will still receive the same care between intervention group, control group, and non-study participants. In fact, the physiotherapists involved in the study would not necessarily even need to be informed of which patients are participating. The control group will receive the information and physiotherapy treatment that is normal practice. Both groups will receive therapy from the range of standard physiotherapy techniques available in the department.

The comparison is thus between patients who are more aware of the availability of treatment choice and those who are offered a normal consultation. The design assumes that the different physical therapies themselves are of similar effectiveness, and also that in the end most patients with similar conditions will receive the same treatment recommendations. Changing this assumption helps to ensure the validity of the study. Thus the main differences in outcomes would arise from the processes of allocating and administering the treatments.

These are conventionally regarded as confounding factors but are the main effects to be measured in this trial. It is likely that the process of informed consent to a conventional trial design would alter patients’ behaviour sufficiently to invalidate the results of this research. The control group may exercise more choice than would otherwise be the case or those with strong preferences may decline to enter the trial. Either would lead to a low and invalid estimate of the intervention’s true effect.

It is proposed therefore to use a modified form of consent. It is proposed that patients be invited to participate in the study by means of an explanatory letter accompanying the initial assessment questionnaires. The letter would also include a consent form to be signed and returned with the questionnaire. If the consent form is not returned, the additional informative pamphlet is not sent out. This change increases the level of informed consent of the participants and reduces the chance of any change in their behaviour or treatment as a result of the study.

Those returning questionnaires and consent forms would then be randomly allocated to the extra information and control groups. However, patients would not be informed of the randomisation procedure and would not be told of the comparison group. All patients will be asked their consent to participate in a research study before they are entered. Any patient who declines to be a part of the study is not given the study materials, nor is their treatment programme affected in any way. In this way, there is no risk of harm to participants who decline the study or who leave the study during treatment.

The risk of harm attached to information giving is most likely very small. No new or different therapy is being offered. Sample group(s) (inclusion criteria): Patients to be included are all people, regardless of gender or ethnic origin, between 18 and 65 years referred by a general practitioner or hospital consultant with a diagnosis of mechanical low back pain with or without leg pain and/or neurological signs. Sample size (itemise numbers for control and treatment arms of the study where applicable): 125 in control arm 125 in intervention arm Excluded persons (enter a description of any types of person explicitly excluded from the study, and explain why): Those unable to attend for outpatient treatment or understand written English.

Or if: any physical therapy has been given in the last three months; recent back operation performed; litigation over back injury in progress or diagnosis of other cause of back pain (effect is likely to be different in all these groups leading to an imprecise estimate of effects and inadequate sample size for sub groups). How was the sample size determined? The minimum effect size that it is desired to detect is a 4 point difference in the ODQ scale. The standard deviation for the within person difference in ODQ scores is expected to be 8 points.

N = {4* [(z alpha = 0.05) + (z beta = 0.2)]} / [(4/8)^2] = 127 To take account of a 10% crossover and 80% follow up doubles this to a total of 250 subjects. What level of significance in sample size calculations have you used? Alpha error of 5% and Beta error of 20% Was there formal statistical input to the overall study design? Yes If yes, please specify the name of the advisor: ----- And the advising department or institution: ----- What events or measures are to be the main outcome measure(s) for the study?

There are no laboratory tests or imaging techniques for assessing response to therapy for back pain. The primary outcome measure to be used for this trial is the Oswestry Back Pain Disability Questionnaire (ODQ). The Quebec back pain disability questionnaire is a relatively new instrument that shows promising psychometric properties. It will be used alongside the ODQ and comparisons of their performance made. As secondary outcomes patients will be asked to rate their satisfaction with their care, and physiotherapists and patients will be asked to rate their judged degree of improvement at the end of therapy.

Strength of belief in different therapies will be measured using a visual analogue or Likert scale. The Internal Control Index questionnaire (a health locus of control measure) will be used to explore if this personality trait modifies the effect of the intervention. All the outcomes are collected by self completed questionnaires sent and returned by post. Patients will be reminded by post and telephone or personal visit up to four times or until they state that they wish to withdraw from the study.

Has this or a related application been submitted elsewhere now or over the past year? Yes If yes, to which organisation? ----- What response have you had/when is a decision expected? ----- Is partial funding being sought from another source? No Have any of the applicants on this form been awarded an NHS grant in the last three years? No Is the research likely to lead to patentable or otherwise commercially exploitable results? If yes, please give details: Yes. It is possible that the information booklet on physiotherapy may be of sufficient quality to be sold.

If this were the case I would seek advice from xxxxxxxx about sale of the copyright and negotiation of royalties for interested parties. (However this is unlikely to amount to very much for anyone I would have thought). Please indicate what methods you plan to use for dissemination or implementation of results: Publication, presentations, commercial partnership Please give details of your dissemination/implementation plans: The research method and findings will be submitted for publication in a high impact peer review journal.

The findings will be presented at relevant conferences and professional meetings and made known through local health service contacts. If it is shown that patient information and choice has a significant health impact then I will explore the possibilities of making this physiotherapy information more widely available through a commercial partnership with a publishing company. Additionally, if the research is found to have major ethical considerations, I will submit the results to publication in ethical journals as well as medical ones.

This change has been made due to the notation in the original review that the ethical possibilities of this research may be even greater than the medical ones. Are any experiments involving animals proposed in your research? No SECTION E: FINANCE AND COSTS Research staff: Research assistant (0.8 WTE): annual salary ?10552; superannuation etc. ?2267; total annual cost ?12819. Costs in year 1: ?12819; in year 2 ?6626; total = ?19445. Travel and subsistence: Year 1 ?800; year 2, ?700; total = ?1500.

Consumables: Stationery and printing: year 1 ?800; year 2 ?400; total = ?1200. Telephone: year 1 ?300; year 2 ?150; total = ?450. Postage: year 1 ?1000; year 2 ?300; total = ?1300. Information booklet writing and printing: year 1, ?1500. Total = ?1500. Copies of ‘the Back Book’ (HMSO) (125 copies): year 1 ?157. Total = ?157.