Options for evidence in Aboriginal communities : Evaluation of therapy for renal disease
Authors : Philip Baker, Peter Morris, Amanda Leach, Wendy Hoy. Menzies School of Health Research, Darwin NT Australia.
Background: There is RCT evidence in diabetic and non diabetic renal disease that angiotension converting enzyme inhibitors (ACEi) have a specific renal sparing effect in addition to their systemic vasoactive effect. However, assessing efficacy of such evidence based medicine is particularity challenging in a setting of isolation, poor compliance, socioeconomic disadvantage, serious alcohol abuse, substandard living conditions, poorly developed health education and deficit chronic disease management. It becomes essential to incorporate "The Measurement Iterative Loop" (Tugwell et al) to evaluate efficiency, monitor the intervention, and re-asses the situation of the full circle.
Setting: In a survey of an Australian Aboriginal community of 1,100 people with a 60-fold elevation in renal failure rates, 30% of adults (18+years) were overweight (BMI>25kg/m2), 18% were diabetic, 11% had impaired glucose tolerance, 26% were hypertensive (BP>140/90), 25% had microalbuminuria (urine albumin/ creatinine ratio (ACR) 3.4 to 34 gm/mole), and 30% had overt albuminuria (ACR>34).
Methods: A five-year treatment program with the long-lasting ACEi perindopril, (Coversyl, Servier Labs) to retard the progression of renal and cardiovascular disease has been initiated. In this prospective community treatment program, process and outcome measures are systematic, compliance evaluated, costs calculated and adverse events evaluated.
Discussion: This presentation will focus on challenges specific to Aboriginal populations and the relevance of the Measurement Iterative Loop. Difficulties of compliance are reduced by the long-lasting, once-a-day dosing of perindopril. Limitation to primarily one ACEi agent provides uniformity for the community. Partnership with community councils and health workers ensures the program is culturally appropriate. Visit attendance is facilitated using "target dates" generated by computerised scheduling. Compliance is assessed by pill counts, specific questioning at each visit and biannual assessment of ACE levels. This study will provide information for the "Implications for Practice", allowing for a broader range of populations to contribute to the Cochrane Initiative.
A domain independent structured trial register
Authors: Nigel Strang, Jean Pierre Boissel French Cochrane Centre. Lyons. France
Background: The Cochrane collaborative review groups (CRG's) build local specialised registers. The Cochrane collaboration also maintains a global bibliographic register that feeds into Medline. Since the early 80's clinical trialists have lobbied for prospective clinical trial registration. This has led to various initiatives with mixed success. Prospective trial registration is the solution of choice to the trial identification problem. Structured trial report information extraction standardises trial report interpretation and the structure increases subsequent search precision. Meta analysis protocols can be partially expressed and standardised as objective search strategies. Geographically dispersed collaborative work is enabled by a shared, centralised resource.
Objectives: Our objective is to explore the nature of and develop an information system exploiting a domain independent structured trial register database that is integrated into the Cochrane process.
Methods: Expert trialists' and meta analysts' knowledge was elicited. Protocol writing Standard Operating Procedures, structured abstracts and trial reports, and existing registers were analysed. Reasons for the mixed success of registers to date were studied.
Results: The structure of the Cochrane collaboration is adapted to the supply of information to a central system. A database structure and information coding mechanism and an internet client server solution, based on the prototype software developed in the TriSum project is proposed. Methods to take into account the justifiable reticence of the pharmaceutical industry are proposed.
Conclusions: Registration should be both retrospective (hand searching and CRG contributions) and prospective, incorporating trials from their inception. Collaborative working and a mechanism for action by Cochrane fields could be enabled by the system. The system could provide a powerful resource for reviewers, consumers, practitioners, administrators and researchers both within and beyond the Cochrane Collaboration. Our proposals are the basis for discussion.
How blind are we?
Authors: García Puig J, Mateos A, Gil A, Barcina C, Bonfill X.
Background: At times the specific assignment of a patient in a "double blind" trial can be disclosed, which may influence the final results of the study.
Objectives: To evaluate the level of knowledge that investigators have about the treatment that each patient receives.
Methods: In the international clinical trial BRILLIANT (Blood pressure, Renal effects, Insulin control, Lipids Lisinopril And Nifedipine Trial), that attempts to compare the effects of lisinopril with those of nifedipine retard in hypertensive patients with diabetes mellitus type II and microalbuminuria over a period of 12 months, the medications are administered according to a classic double blind concealment procedure. During the study a basal level of the excretion of microalbuminuria is determined, and again at 6 and 12 months. A clinical evaluation is made at each visit.
Upon completion of the study, and before being informed of the real aleatorial assignment, four of the participating investigators tried to predict the medication that a patient received based on the adverse reactions detected and the modifications of renal excretion of albumin based on pre-established criteria. A total of 33 patients were analyzed, and followed actively for 12 months, after which the aleatorization code was revealed.
Results: In 16 patients treated with lisinopril there was agreement with the predictions in 12-14 cases (75-87.5%). Of the 17 patients treated with nifedipine, it was predicted that 15 were treated with this drug. These results indicate that the double blind could have been revealed in 27-29 of the 33 randomized patients (82-88%).
Conclusions: In order to improve the double blind method, one alternative is prevent investigators from knowing the clinical data relevant to the trial -in this case the microalbuminuria levels. Another alternative would be to separate the clinical management of the patient from the evaluation of the treatments.
Setting up a Field. Can strategic management help?
Authors: EJ Dickinson, R Dickinson (Royal College of Physicians, UK).
Background: Fields are the new development in the Collaboration but only four have registered. Perhaps Field establishment is a difficult process about which little is known.
Objective: To report learning from the initiation of a Field (Health Care of Older People).
Methods: Qualitative study using a basic strategic management model
Results: Field establishment involves iteration between analysis (four issues) and implementation (three issues).
Analysis:
Implementation:
Conclusion: Establishing Fields is complex and can be successful by utilising aspects of basic strategic management.
Adding value to hand searching
Authors: Mark Lodge, David Gill
Background: Electronic databases provide unreliable access to identifying randomized controlled trials (RCTs). Hand searching journals remains a costly necessity. We can add to the value of hand searching by conducting epidemiological studies of the trials we find. We report a study, assessing the accessibility and methodological quality of the RCTs found in the British Journal of Surgery (Br J Surg).
Objective: To asess the profile of RCTs found by handsearching Br J Surg 1947-1993.
Methods: The Br J Surg was hand searched in 1994 and each RCT coded as to whether:
Results: 552 RCTs were identified by handsearching, of which only 83 had been tagged on MEDLINE PT=RCT in 1994, although 426 had used the term "random" in the title or abstract. Of the 552, 417 were full length articles. Of these 417, 179 included some description of the randomization procedure, 191 accounted for all the subjects in the analyses and 138 reported that the outcomes had been blindly assessed.
Conclusions: Cochrane entities, such as Fields, are ideally placed to undertake reviews of key journals, identifying those areas where the quality of reported research is poor and indicating ways in which journals can be more effective in making trials identifiable for indexers and reviewers.
Is it Necessary to Treat Acute Otitis Media for 10 Days? A Meta-Analysis
Authors: Anita Kozyrskyj, Elske Hildes-Ripstein, Sally Longstaffe, Leigh Wincott, Daniel Sitar, Terry Klassen, Michael Moffatt
Background: Acute otitis media (AOM) is a common childhood illness -which frequently results in a prescription for an antibiotic. In North America, the Standard treatment of AOM is a 10-day course of antibiotics- However, a shortened course of antibiotics may decrease costs and the risk of developing resistant bacteria.
Objectives: To conduct a meta-analysis of RCTs of the antibiotic treatment of AOM in children, with the primary objective of determining whether clinical resolution following treatment with antibiotics for <7 days was equivalent to treatment for> 7 days.
Methods: Trials of the antibiotic treatment of AOM were identified in the medical literature and assessed for inclusion according to the following criteria: I month to 18 years of age, clinical diagnosis of AOM, and randomization to treatment with <7 days of antibiotics or> 7 days. The methodologic quality of included trials was assessed (Jadad scale) by 7 reviewers. Outcome data, documented as the number of treatment failures, were extracted from individual trials, and the cumulative odds ratio and risk difference were calculated using a fixed effects model. Sensitivity analyses were conducted to assess the robustness of the meta-analysis.
Results: Eighteen trials (mean quality score = 2.7, SD= 1 -0) met inclusion criteria and were grouped by antibiotic used in the short treatment arm (1. shortacting antibiotics, eg cefaclor 2.ceftriaxone 3. azithromycin). The _<1 month outcome (OR="1.17," 95% CI-0.89-1.53) and < 3 month outcomes (OR="0.9," 95% CI="0.59-1.37)" of treatment with 5 days of short-acting antibiotics were not significantly different from treatment with>7 days No differences in outcome between treatment arms were shown for the ceftriaxone and azithromycin groups. Sensitivity analyses confirmed the robustness of the meta-analysis of the short-acting antibiotic group.
Conclusion: The results of this meta-analysis show that 5 days of short-acting antibiotic is effective treatment for AOM in children.
Title : Randomized controlled trials (RCTs) in certain collagen vascular diseases
Author: Yoichi Nagayama
Background: Early diagnosis and treatment, by conventional methods, of patients with systemic lupus (SLE), systemic sclerosis (pss) and poly/dermato myositis (PDM) remarkably improve prognosis resulting in an increase in number of chronic patients in whom underlying abnormalities persist. Development of internal organ associations leading to life-threatening conditions such as lung disease has become increasingly common.
Objectives: To review RCTS of novel effective therapy for untoward serious manifestations in these Patients.
Methods: Electronic searching of MEDLINE between 1963 and 1996.
Results: Two reports on SLE were found; combination therapy and hemapheresis for brain and kidney lesions, two reports on pss; prostacyclin analogues for lung disease and cisapride for esophageal lesions. No report on PDM.
Conclusions: There were few RCTs. Those were short term frame and involved insufficient number of patients. Therefore, further studies are needed.
WHO-ISH (World Health Organisation and International Society of Hypertension) Blood Pressure Lowering Treatment Trialists' Collaboration: a register of trials for the conduct of overviews
Authors: B Neal and S MacMahon on behalf of the WHO-ISH Collaboration. Clinical Trials Research Unit, University of Auckland, New Zealand
Background: A number of randomised trials have been initiated to address the current uncertainty surrounding the efficacy of newer blood pressure lowering agents. Many of these may not be individually large enough to detect plausibly modest differences between treatments.
Aims:
Methods: A register of major ongoing and planned randomised trials has been established and collaboration sought from the principal investigators of each study. A detailed protocol has been prepared for overviews of individual patient data with anayses scheduled for 1999 and 2002.
Results: 29 eligible trials have been identified and characterised. These trials will provide a total of approximately 1,000,000 patient years of follow-up from 200,000 randomised patients. It is anticipated that more than 5,000 strokes and 10,000 coronary heart disease events will be documented.
Conclusions: The overviews planned should have good power to resolve the principal persisting uncertainties about the relative efficacy of newer and older agents. The prospective design and the use of individual patient records will overcome the problems associated with retrospective overviews based on tabular data.
The Power of Placebos
Author: Mark Turner (Department of Orthopaedic Surgery and Musculoskeletal Medicine, Christchurch School of Medicine, University of Otago, New Zealand)
Background: While the reporting of Randomized Controlled Trials (RCT's) is being standardised and improved, the role of the placebo effect is often not considered when reporting the results of these trials. By not taking account of the power of placebo effects authors and readers risk misinterpreting the results of RCT's.
Objectives: Examples of the power of placebo effects will be presented and several theories of placebo mechanisms will be examined. Proposed theories of placebo action include conditioning, faith, expectation, design and measurement artefacts, anxiety reduction, homeostatic-cybernetic self regulation, and symbolic processes (White, Tursky & Schwartz, 1985). The implications of the conditioning theory of placebo action will be examined in detail. By understanding how placebo effects can be generated it is possible to better take account of these factors. The conditioning model of pharmacotherapy suggests there may be a 'psychological carryover effect' which is distinct from any pharmacokinetic influence when active medication is presented first (Ader, 1993).
Methods: Examples of RCT's that have inadvertently maximised the placebo effect will be presented, along with possible explanations from differing theoretical perspectives, in order to demonstrate how unjustifiable conclusions can be drawn from the results of RCT's when the power of placebos is not adequately considered.
Conclusion: While placebo effects should no longer be considered 'nuisance variables', an understanding of how they can be generated makes it possible to better take account of their potential effects.
References:
Ader, R. (1993). Conditioned responses in pharmacotherapy research. Psychological Medicine, 23, 297-299.
White, L., Tursky, B., & Schwartz, G.E. (eds.). (1985). Placebo: Theory research and mechanisms. New York: The Guilford Press.
Informed Consent in Systematic Reviews
Author: Dr Paul White Clinical Studies Unit, Australia
Background: While there are many publications reviewing the principles of informed consent (IC), there are little data to indicate if the absence of informed consent affects the outcome of Randomised Controlled Trials (RCT's). This study aimed to assess RCT's in The Australian and New Zealand Journal of Psychiatry (ANZJP) with regard to explicit reporting of informed consent (IC).
Issues related to inclusion of studies that have not reported that IC was gained are discussed. Some clinical trials are still undertaken where there is doubt about the quality of the informed consent obtained. This neglect may introduce biases into systematic reviews. There is little empirical evidence to support this. However studies are suggestive and indicate the need for further research.
Methods: The ANZCP was hand searched for RCT's from 1967 -1995. All RCT's were inspected in order to determine whether IC was reported. 43 RCT's were identified. 10 reported that informed consent was obtained and 33 did not. There was a statistically significant improvement in reporting over time.
Conclusions: IC is poorly reported in the ANZJP. Comparative studies from other psychiatric journals show even poorer results. The necessity of obtaining informed consent introduces a potential sampling bias. Only those patients who can and are willing to consent are included in RCTs. Evidence suggests that this may be a more compliant group than those who do not enrol. These diverse groups potentially introduce a source of heterogeneity into attempts to combine RCTs in which consent was obtained and in which it was not. There is a moral argument to exclude RCT's from systematic reviews which do not report IC. If consent was not obtained from patients involved in RCTs should that data be used in systematic reviews i.e. is science which is only potentially unethical "bad science"?
Journal of Clinical Oncology (JCO): How systematic are reviews?
Authors: Vivien Bramwell, Chris Williams
Background: High quality reviews are a critical resource for the clinician. Mulrow (Ann. Int. Med. 1987; 106:485) showed reviews in 4 medical journals did not routinely use scientific methods to identify, assess and synthesise data.
Objective: To perform a qualitative analysis of reviews in JCO.
Methods: Hand search of JCO, Jan 1983-Dec 1995, for reviews and meta-analyses of cancer treatment. Quality was assessed using 8 criteria proposed by Mulrow, rated independently by 2 medical oncologists as: specified (S), unclear (U), not specified (N).
Results: 122 (69%) of 176 reviews dealt with treatment aspects. Meta-analyses (total 12) appeared from 1991, structured abstracts from 1992.
Reviews (106) Meta-Analyses (16)
Criteria: S U N S U N
Purpose 105 1 0 16 0 0
Data Identification 12 11 83 11 2 3
Data Selection 11 21 74 13 2 1
Validity Assessment 9 15 82 9 5 2
Qual. Synthesis 106 0 0 16 0 0
Quan. Synthesis 1 1 104 16 0 0
Summary 101 4 1 15 1 0
Future Directions 81 15 10 15 1 0
Authors rarely gave information on literature searches, criteria for study inclusion and quality assessment. Quantitative summation was confined to meta-analysis. Results analysed by publication year show recent improvement, especially since 1991, consequent upon inclusion of meta-analyses and, possibly structured abstracts.
Conclusions: Reviews of cancer treatment in JCO were not systematic, and methodology needs improvements. Meta-analyses proved an exception. The Cochrane Cancer Network aims to improve the scientific rigour of clinical reviews. It will also prepare, maintain and disseminate systematic reviews of randomised clinical trials in cancer.
This page was last updated on 18 November 1996
Comments for improvement or correction are welcome.
georg.koch@cochrane.de