|
Randomised controlled trial of near-patient testing in people with type 2 diabetes mellitus |
Khunti K, Stone M A, Burden A C, Turner D, Raymond N T, Burden M, Baker R |
|
|
Record Status This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn. Health technology The study compared two diagnostic procedures for testing glycated haemoglobin (HbA1c) in patients with Type 2 diabetes. The diagnostics tests compared were near-patient (rapid) testing and normal laboratory testing for HbA1c. Near-patient (rapid) testing is a technology offering on-site testing of patients for HbA1c (the equipment used in the current study was the Bayer DCA 2000, which requires a finger prick instead of formal venesection and provides the results in 6 minutes).
Economic study type Cost-effectiveness analysis.
Study population The study comprised patients with Type 2 diabetes who attended general practices to manage their condition. Patients who could not attend the practice, or who received only hospital care, were excluded from the study. No further inclusion or exclusion criteria were reported.
Setting The setting was primary care. The economic study was carried out in Leicestershire, UK.
Dates to which data relate The patients were recruited into the study between August 2000 and May 2002. The resource use data were recorded during the follow-up period. The cost data related to 2002 to 2006. The price year was 2002/03.
Source of effectiveness data The effectiveness data were derived from a single study.
Link between effectiveness and cost data The costing was carried out prospectively on the same sample of patients as that used in the effectiveness study.
Study sample Power calculations demonstrated that, with nearly 300 patients in each group and an increase of 10% in the proportion of intervention group patients (near-patient test), the power of the study would be 70%. However, if the intervention could result in a 12% increase in good control, the power of the study would increase to 80%. Nurses recruited the patients. A total of 681 patients (90% of invited participants) were recruited into the study from 8 general practices. Of these, 343 were randomly allocated to the near-patient test group (intervention group) and 338 to the normal test group (control group). Twenty-four patients were subsequently excluded from the intervention group, 4 because they had Type 1 diabetes, 2 because of impaired glucose tolerance (IGT), and 18 because they had no further HbA1c results after recruitment. Nineteen patients were excluded from the control group, 2 due to poor general health, 6 because they had Type 1 diabetes, and 11 because had no further HbA1c results after recruitment. The final sample therefore comprised 638 patients (44 to 115 per practice).
Study design The analysis was based on a multi-centre, prospective, randomised controlled trial. Practising nurses carried out the randomisation at patient level within participating general practices using numbered sealed envelopes. The envelopes followed a computer allocated block randomisation in blocks of four. The allocation sequence was defined by the research team. Each patient was followed up for 12 months after recruitment. At the end of the follow-up period, 9% of the initial sample was lost to follow-up in the intervention group, compared with 3.1% in the control group. Patients were lost to follow-up because they left the practice, died, or failed to attend appointments. It was reported that neither the health professionals nor researchers were blinded to the intervention.
Analysis of effectiveness The analysis was conducted on an intention to treat basis, as well as on protocol samples (i.e. patients with available laboratory results at baseline and at 12 months' follow-up). It was reported that the patient groups were comparable at analysis. The primary health outcome was the proportion of patients with "good glycaemic control", defined as an HbA1c of less than 7%. To facilitate comparisons, laboratory tests were taken for all patients at baseline and at 12 months at one laboratory (Bio-Rad Variant II high-performance liquid chromatography). To adjust for potential confounding factors and possible biases, the authors carried out multiple logistic regression analysis using "good glycaemic control at follow-up" as the dependent variable. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated for intervention patients compared with controls. In the analysis, the authors controlled for the effect of gender, general practice attended, duration of diabetes, treatment at baseline, age, deprivation variables and baseline glycaemic control status.
Effectiveness results The proportion of patients with good metabolic control (HbA1c <7.0%) was 0.37 (95% CI: 0.32 to 0.42) in the intervention group versus 0.38 (95% CI: 0.33 to 0.43) in the control group. The difference was not statistically significant.
The unadjusted OR for intervention patients versus controls with good glycaemic control at follow-up was 0.95 (95% CI: 0.69 to 1.31).
The OR for the intervention versus control group was 0.80 (95% CI: 0.56 to 1.15) when adjusted for baseline HbA1c status.
The multiple logistic regression analysis demonstrated the best fit to the data.
The model included baseline HbA1c status, gender, duration of diabetes treatment at baseline and general practice attended (OR 0.84, 95% CI: 0.58 to 1.22).
If the analysis was conducted only for patients with complete protocols (n=599), the unadjusted OR was 0.97 (95% CI: 0.70 to 1.36) and the adjusted OR was 0.88 (95% CI: 0.60 to 1.29).
Clinical conclusions The analysis demonstrated that the use of the near-patient test to measure HbA1c in patients with Type 2 diabetes in primary care does not improve metabolic control in comparison with normal laboratory testing for HbA1c.
Measure of benefits used in the economic analysis The authors did not derive a summary measure of benefit in the economic analysis. In effect, the study was a cost-consequences analysis.
Direct costs The health service costs included in the analysis covered diabetic medication, consumables (including blood and urine strips, lancets and sundries), general practice contacts (including contacts with a nurse, general practitioner (GP), phlebotomist or chiropodist at surgery), GP or nurse home visits, GP or nurse telephone consultations, and hospital inpatient and outpatient services. In addition, health service-related patient-borne costs were also included. Such costs were travel costs (including mode of travel), mileage, fares and parking fees. A trained data collector recorded resource use for prescribing, consultations, and use of urine and blood test strips, lancets and sundries (i.e. needles) and secondary services from GP records using standardised data collection forms. The duration of telephone consultations was based on authors' assumptions and published data. The quantities for travel costs were obtained directly from the patients (83% of patients included in the intention to treat analysis) who completed a specifically designed questionnaire. Data on health service costs and travel costs were obtained from official published sources or from the authors' setting (University Hospitals of Leicester NHS Trust).
The costs and the quantities were not reported separately, although the cost per patient in each category was reported. Discounting was not relevant as the costs were incurred during less than 2 years. With the exception of the HbA1c test, which was reported for the price year 2003, all costs were reported for the price year 2002/03.
Statistical analysis of costs The costs were treated deterministically.
Indirect Costs The authors accounted for productivity losses, defined as time spent for travelling and for receiving care. A different value was assigned to working and non-working time. Relevant quantities of resources were based on actual patient data, derived by means of specifically designed questionnaires. The cost data were derived from official published sources. All data were reported for the price year 2002/03. Discounting was not relevant as the costs were incurred during less than 2 years.
Sensitivity analysis No sensitivity analysis was carried out.
Estimated benefits used in the economic analysis See the 'Effectiveness Results' section.
Cost results The total health service related costs were 122,718 in the control group (389.58 per patient) and 116,696 in the intervention group (370.46 per patient). The cost-difference between the two groups was not statistically significant.
The total patient-borne costs were reported per patient and were 52.47 in the control group and 50.31 in the intervention group.
There was no statistically significant difference between the two groups.
Synthesis of costs and benefits The costs and benefits were not combined.
Authors' conclusions "Near-patient testing for HbA1c (glycated haemoglobin) alone does not lead to outcome or cost benefits in managing people with Type 2 diabetes in primary care."
CRD COMMENTARY - Selection of comparators The selection of the comparators was explicitly justified. Normal laboratory testing would seem to represent standard practice in the authors' setting.
Validity of estimate of measure of effectiveness The analysis was based on a multi-centre, prospective, randomised controlled trial, which was appropriate given the study question. The study sample was representative of the study population and the patient groups were shown to be comparable at analysis. In addition, participating general practices were representative since they covered urban and rural as well as deprived and affluent areas. The study methodology (sample size determined by power calculations, method of randomisation, duration of the study, and losses to follow-up) should ensure a high internal validity for the study. In addition, statistical analysis was conducted to account for potential biases and confounding factors. However, owing to certain limitations, the power of the study (as defined by the power calculations) was less than 80%.
Validity of estimate of measure of benefit The authors did not derive a summary measure of benefit in the economic analysis. The analysis, in effect, was a cost-consequences analysis. This approach, although clearly valid for the intervention, does not permit comparisons with other health care programmes as would be the case with a cost-utility analysis.
Validity of estimate of costs The perspective adopted in economic analysis was not explicitly reported. However, it might have been a societal perspective since both indirect and health service cost categories were included in the analysis. In the estimation of patient-borne costs, time and productivity losses due to home visits and telephone costs were omitted. It is unlikely though that their omission would have affected the authors' conclusions. Summary costs were reported for each cost category but the costs and the quantities were not reported separately, which will present difficulties when reworking the analysis in other settings. Resource use was based on actual data (GP records and patient questionnaires), but no sensitivity analysis was conducted to assess the robustness of the estimates used. The cost data were derived from official published sources and the authors' setting. The cost data were treated deterministically and no sensitivity analysis was conducted, which might introduce some uncertainty into the results. Appropriate adjustments for inflation and the price year were, however, reported.
Other issues The authors compared their study findings with those of other studies and attributed differences in the results to differences in methodology, setting and baseline levels of glycaemic control. The issue of generalisability of the results to other settings was not directly addressed. The main findings of the study and the results of the statistical tests were thoroughly reported. The study considered patients with Type 2 diabetes mellitus and this was reflected in the authors' conclusions.
The authors reported a number of limitations to their study. First, owing to funding constraints, randomisation had to be conducted at patient level instead of practice level. Consequently, the patients' follow-up was not monitored closely and did not allow differences in GP visits to be demonstrated as, in many cases, patients continued to attend more than once. In addition, neither of the participating health professionals was blinded to the intervention, and it is therefore possible that they may have intervened with the form of care provided to patients.
Implications of the study The authors recommend that "general practices should not rush to adopt the use of in-house testing for HbA1c on the assumption that this alone will lead to better glycaemic control within their patient population". However, given the limitations and constraints of their study, they recommended that future research should be carried out to evaluate rapid tests as part of diabetes patient management.
Source of funding Funded by the NHS Executive (Trent), UK.
Bibliographic details Khunti K, Stone M A, Burden A C, Turner D, Raymond N T, Burden M, Baker R. Randomised controlled trial of near-patient testing in people with type 2 diabetes mellitus. British Journal of General Practice 2006; 56: 511-517 Other publications of related interest Because readers are likely to encounter and assess individual publications, NHS EED abstracts reflect the original publication as it is written, as a stand-alone paper. Where NHS EED abstractors are able to identify positively that a publication is significantly linked to or informed by other publications, these will be referenced in the text of the abstract and their bibliographic details recorded here for information
Indexing Status Subject indexing assigned by NLM MeSH Aged; Blood Glucose /metabolism; Costs and Cost Analysis; Diabetes Mellitus, Type 2 /blood /economics; Family Practice /economics /organization & Female; Hemoglobin A, Glycosylated /analysis; Humans; Hyperglycemia /economics /prevention & Hypoglycemia /economics /prevention & Male; Point-of-Care Systems /economics; Prospective Studies; administration; control; control AccessionNumber 22006008278 Date bibliographic record published 28/02/2007 Date abstract record published 28/02/2007 |
|
|
|