|Home-based management can achieve intensification cost-effectively in Type I diabetes
|Dougherty G, Schiffrin A, White D, Soderstrom L, Sufrategui M
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.
Home-based management for children with newly diagnosed insulin-dependent diabetic mellitus (IDDM). The key aspect of the intervention programme was the availability of a diabetes-treatment nurse who accompanied the family to the home and offered flexibly scheduled teaching sessions and who, in collaboration with the attending diabetologist, implemented initial and almost all subsequent insulin treatment.
Economic study type
Children with IDDM presenting to the emergency department of a children hospital, who fulfilled the following inclusion criteria: over 2 years of age, no sibling with IDDM, living at home, and living within 1 hour of the hospital.
Children's hospital (secondary and tertiary-care paediatric hospital) and community. The economic study was carried out in Canada.
Dates to which data relate
Effectiveness and resource use data corresponded to patients recruited for the study between October 1989 and September 1991. The price year was 1991.
Source of effectiveness data
The evidence for the final clinical outcomes was derived from a single study.
Link between effectiveness and cost data
Costing was retrospectively undertaken on the same patient sample as that used in the effectiveness analysis.
Power calculations were used to determine the sample size (the study was set to detect a difference in the mean 2nd-year glycosylated haemoglobin (HbA1c) level of 1.5% points with 95% power). The study sample consisted of 63 (out of 73) children with IDDM, who fulfilled the inclusion criteria, and who were randomly assigned to either the home-based group (n=32, mean (SD) age 10.7 (3.9) years) or to the hospital-based group (n=31, mean (SD) age 9.8 (3.9) years). The participation rate was 100%. The treating team for both groups consisted of 3 diabetologists, a psychologist, and a social worker.
The study was a randomised controlled trial, carried out in a single centre. The treatment period was for 2 years after diagnosis and patients were followed for 1 year after the end of the intervention programme. Stratification of patients was performed in terms of age and sex. Loss to follow-up was not explicitly reported. The patients, parents, and treating team were blind to the aggregate group data about patient characteristics, insulin treatment, and clinical or psychological results during the study period. Laboratory technicians assessing HbA1c results were kept blind to the patients' treatment assignment. An independent evaluation team was responsible for all data collection and analysis.
Analysis of effectiveness
The principle used in the analysis of effectiveness appears to have been treatment completers only. The health outcomes were HbA1c concentration as a measure of intermediate-term glycemic control (as measured quarterly during the first 24 months and then at 36 months), diabetes-related adverse events, knowledge of diabetes (as assessed with the Diabetes Knowledge Scale (DKN) plus five more items added as a result of a ceiling effect observed during pretesting), adherence to the treatment (as evaluated by the Diabetes Regimen Adherence Questionnaire-R (DRAQ), a 20-item, self-report measure), family impact (as measured by the Impact on Family Scale and the Perceived Stress Scale), satisfaction (as assessed with a 10-item questionnaire developed for this study), child behaviour (as measured by the Achenbach Child Behavior Checklist (CBCL)), life events using a modified life events scale which was completed by both parents and the treating team (scores were counts of life events), and absence from school. It was found that a higher proportion of subjects in the home-based group were Tanner Stage III-IV at diagnosis than in the hospital group (59% versus 32%), representing a disadvantage for the home-based arm of the study. Linear regression was used to adjust insulin doses for an observed imbalance in Tanner Stage at diagnosis.
The mean (SD) HbA1c values at 24 months were 6.8 (1.3%) for the hospital-based group and 6.1 (1.3%) for the home-based group.
The corresponding values at 36 months were 7.1 (1.3%) for the hospital-based group and 6.4 (1.4%) for the home-based group, (F = 5.55; p<0.02).
The number of diabetes-related adverse events was comparable in the two study groups. No significant differences were observed in terms of knowledge, regimen adherence, family impact, satisfaction, life events, and absence from school.
The hospital-based group scored significantly lower (p<0.05) in terms of the Perceived Stress Scale at 1 month.
Superior glycemic control was observed in the home-based arm during the 2nd year of follow-up and for 1 year after the intervention ended. This was associated with a similar rate of diabetes-related adverse events in the two groups.
Measure of benefits used in the economic analysis
No summary benefit measure was identified in the economic analysis, and only separate clinical outcomes were reported separately.
Costs were discounted. Some quantities were reported separately from the costs. Cost items were not reported separately. Cost analysis covered 3 categories of costs:
(1) health care services (diagnostic, therapeutic, medical, and hotel services) using data from patients' hospital records,
(2) diabetes-treatment nursing as reported by nurses themselves; and
(3) parents' diabetes-related out-of-pocket spending (baby-sitters, travel, hospital meals, etc) recorded using a structured questionnaire completed every 3 days by telephone for the 32 days after diagnosis.
The perspective adopted in the cost analysis was that of society. The price year was 1991. The cost analysis did not include the hospital overhead costs as the related resource use was deemed to be unaltered by use of the home-based programme.
Statistical analysis of costs
The Student t test was used to compare costs between the groups.
Costs were discounted. Some quantities were reported separately from the costs. Cost items (average value of parents' time based on average hourly earnings of men and women) were reported separately. Cost analysis covered the indirect costs associated with work and nonwork time of parents; since the average time off work was similar in both groups, the valuation was based only on nonwork time. The perspective adopted in the cost analysis was that of society. The price year was 1991.
A series of one-way sensitivity analyses was performed on cost data.
Estimated benefits used in the economic analysis
The discount rate was 5%. The mean total cost per case was only Can$48 higher in the home-based programme (NS) compared with the hospital based programme. The home-based intervention was associated with significantly lower out-of-pocket expenses and higher health care costs.
Synthesis of costs and benefits
Costs and benefits were not combined since the intervention was the weakly dominant strategy (equivalent social costs and better clinical outcomes). The cost estimate was found to be sensitive to the assumptions regarding certain key variables of the calculations.
The authors concluded that home-based management for newly diagnosed diabetic children can result in better metabolic control and similar psychological outcomes compared with traditional hospital and clinic-based care, without notable effects on social (total) costs.
CRD COMMENTARY - Selection of comparators
A justification was given for the choice of the comparator (the strategy of hospital-based management): it was the traditional approach used in the context in question. You, as a database user, should consider whether this is a widely used health technology in your own setting.
Validity of estimate of measure of effectiveness
The effectiveness results are likely to be internally valid given the randomised study design, power calculations, high participation rate, the use of an independent evaluation team, the extensive use of standardised measurement instruments, and adjustments made for the effects of differences between the study groups at baseline. The potential bias introduced into the study because different nurses were used in each of the two study groups should, however, be borne in mind; as discussed by the authors. The effectiveness analysis appears to have been based on the principle of treatment completers, introducing some potential bias into the study. The study groups were different in terms of Tanner Stage III-IV at diagnosis, representing a disadvantage for the home-based arm of the study. However, adjustments were made for the effects of this difference. It was deemed that the patient sample was representative of the study population along the socioeconomic status (SES) dimension.
Validity of estimate of measure of benefit
The authors did not derive a measure of health benefit. The economic analysis was therefore of cost-consequences design.
Validity of estimate of costs
Some quantities were reported separately from the costs. Adequate details of methods of cost estimation were given. The effects of different procedures on indirect costs (productivity loss) were addressed. Some statistical analyses were performed on resource use and cost data. The price year was given. The conversion rate from Canadian currency to US dollars was not reported.
The authors' interpretation of the study results appears to be justified given the uncertainties in the data. The issue of generalisability of costs to other settings or countries was addressed by performing sensitivity analysis on costs; regarding effectiveness results, it was deemed that representative age, and sex distributions, high participation rate and representative sample along the dimension of SES were supportive. Some comparisons were made with other studies.
Implications of the study
This study provides evidence that intensification can be achieved cost-effectively in a high proportion of patients with a treatment strategy involving minimal hospitalisation at diagnosis and home-based nursing follow-up. The more conventional approach may still remain the treatment of choice for very young children, for patients or families unable to assume some aspects of diabetes management rapidly after diagnosis, or for families with significant psychological problems, and in centres or regions unable to provide or develop well-organised, specialised, home-based services.
Source of funding
Supported by a grant from the National health Research and Development Program of Canada.
Dougherty G, Schiffrin A, White D, Soderstrom L, Sufrategui M. Home-based management can achieve intensification cost-effectively in Type I diabetes. Pediatrics 1999; 103(1): 122-128
Subject indexing assigned by NLM
Child; Cost of Illness; Cost-Benefit Analysis; Diabetes Mellitus, Type 1 /blood /economics /psychology /therapy; Female; Health Care Costs; Hemoglobin A, Glycosylated; Home Care Services /economics; Hospitalization /economics; Humans; Male; Outpatient Clinics, Hospital /economics
Date bibliographic record published
Date abstract record published