|Multisystemic treatment of poorly controlled type 1 diabetes: effects on medical resource utilization
|Ellis D A, Naar-King S, Frey M, Templin T, Rowland M, Cakan N
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.
The authors studied multisystemic therapy (MST), an intensive, home-based, psychotherapy for the improvement of control of Type 1 diabetes. MST comprised family-centred, community-based treatment designed to target the multiple systems within which youth with serious problems and their families are embedded. The therapy consisted of a multisystemic assessment with treatment goals and interventions tailored to each family to improve adherence to diabetes treatments. Therapists were expected to meet with families a minimum of 2 to 3 times per week at the beginning of treatment. Treatment was expected to last for approximately 6 months, ending when treatment goals were met.
Type of intervention
Secondary prevention of diabetes-related adverse events through the improvement of metabolic control.
Economic study type
The study population comprised adolescents diagnosed with Type 1 diabetes for at least 1 year with glycohaemoglobin (GHb) of 13% or higher. Adolescents were included if they had a Tanner stage of II or above to be certain that they had entered puberty. Adolescents were excluded if they were older than 16 years of age, or if they had moderate or severe mental retardation or psychosis.
The setting was the community, located in a major metropolitan area (not specified). The economic study was carried out in the USA.
Dates to which data relate
The dates during which the effectiveness and resource use data were collected were not clear. The authors referred to a parent study for the effectiveness data, but bibliographic details were not provided.
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 using the same patient sample as that used in the effectiveness study.
The study sample was recruited from an endocrinology clinic within a tertiary care children's hospital. There was no report that power calculations were carried out to estimate the impact of chance on the results. A total of 47 adolescents were found to be eligible. Of these, 9 refused to participate and 7 did not complete baseline data collection, leaving 31 adolescents to be randomised into the study. Sixteen adolescents were allocated to the intervention group and received home-based psychotherapy in addition to their standard medical care. The remaining 15 adolescents were allocated to the control group.
This was a randomised controlled trial. Details of the method of randomisation were not reported. The study was based at a single centre (a tertiary care children's hospital in a major metropolitan area). The patients were followed for 6 months over the course of study treatment, and for a further 3 months to assess maintenance of treatment effectiveness after the treatment had ended. It was reported that 5 adolescents in the intervention group and 3 in the control group were lost to follow-up. The method of blinding was reported.
Analysis of effectiveness
The analysis was conducted on an intention to treat basis. Rates of hospital utilisation were used as a proxy for health outcome. Specifically, the authors measured the mean number of ER visits and hospital admissions for the 9-month period before study entry and for the 9 months after treatment initiation. Metabolic control was also measured using total GHb, and the relationship between changes in inpatient admissions and changes in metabolic control was investigated. The authors reported that there were no significant differences between the MST and control groups in terms of the GHb level or demographical variables at baseline.
The mean number of ER visits at baseline was 0.06 (standard deviation, SD=0.25, range: 0 to 1) for MST and 0.27 (SD 0.46, range: 0 to 1) for the control group.
The mean number of admissions at baseline was 0.56 (SD 1.03, range: 0 to 3) for MST and 0.47 (SD 1.13, range: 0 to 4) for the control group.
The percentage of patients with no admissions at baseline was 68.7% for MST and 80.0% for the control group.
The percentage of patients with a single admission at baseline was 18.8% for MST and 6.7% for the control group.
The percentage of patients with multiple admissions at baseline was 12.5% for MST and 13.3% for the control group.
The mean number of ER visits during the study period was 0.13 (SD 0.34, range: 0 to 1) for MST and 0.47 (SD 0.64, range: 0 to 2) for the control group.
The mean number of admissions during the study period was 0.13 (SD 0.34, range: 0 to 1) for MST and 1.00 (SD 2.14, range: 0 to 7) for the control group.
The percentage of patients with no admissions during the study period was 87.5% for MST and 73.3% for the control group.
The percentage of patients with a single admission during the study period was 12.5% for MST and 6.7% for the control group.
The percentage of patients with multiple admissions during the study period was 0% for MST and 20.0% for the control group.
There was a statistically significant difference in change in number of admissions between the two groups, (p<0.01).
There was no statistical difference in the change in the number of ER visits between the two groups.
For the MST group, participants with a decreasing number of hospital admissions were also found to have decreased GHb and, therefore, improved metabolic control (correlation, r=0.58, p<0.05). This association was not true for the control group (r = -0.15, p non significant).
The authors concluded "ST has the potential to decrease inpatient admissions among adolescents with poorly controlled Type 1 diabetes".
Measure of benefits used in the economic analysis
The authors did not estimate a summary measure of health benefit. The study was, in effect, a cost-consequences analysis.
The costs were estimated from the perspectives of the hospital and the third-party payer. The authors compared costs between two periods (9 months before randomisation and during the 9-month study period after randomisation). The cost categories included were hospital direct costs, hospital charges and insurance costs. No further details on the cost items were reported. Resource use data were collated from the clinical study from the hospital's computerised medical information system. The hospital finance department provided information on charges, costs and revenue. Only visits that were relevant to diabetes were included in the analysis. The authors measured the direct costs associated with ER and inpatient services to inform the hospital perspective and hospital revenues to inform the third-party payer perspective. The costs and the quantities were reported separately. Discounting of the costs was not necessary given the relatively short time period of the analysis (less than 2 years). A price year was not reported.
Statistical analysis of costs
Repeated measures (analysis of variance) were conducted using hospital charges, hospital costs and insurance costs as dependent variables and group assignment as the independent variable.
The indirect costs were not relevant to the perspectives of the study.
There was no report that sensitivity analyses were carried out.
Estimated benefits used in the economic analysis
Not relevant, see the 'Effectiveness Results' section.
The mean hospital charges for MST were $3,455 (SD 7,454, range: 0 to 22,992) at baseline and $1,902 (SD 5,742, range: 0 to 22,531) during the study period.
The mean hospital direct costs for MST were $990 (SD 2,056, range: 0 to 6,188) at baseline and $382 (SD 1,114, range: 0 to 4,325) during the study period.
The mean third-party payer costs for MST were $3,334 (SD 7,426, range: 0 to 28,228) at baseline and $923 (SD 2,503, range: 0 to $8,470) during the study period.
The mean hospital charges for standard care were $3,464 (SD 8,064, range: 0 to 23,531) at baseline and $7,513 (SD 15,415, range: 0 to 46,497) during the study period.
The mean hospital direct costs for standard care were $1,101 (SD 2,623, range: 0 to 8,157) at baseline and $2,038 (SD 4,238, range: 0 to 13,279) during the study period.
The mean third-party payer costs for standard care were $2,740 (SD 6,065, range: 0 to 20,873) at baseline and $3,571 (SD 6,425, range: 0 to 20,578) during the study period.
Charges (f=2.87, p<0.05) and direct hospital costs (f=2.81, p<0.05) for those receiving MST decreased significantly over the study period compared with standard care.
Synthesis of costs and benefits
"MST (multisystemic therapy) can successfully treat adolescents with poorly controlled diabetes by reducing the frequency with which they are admitted to hospital and reducing the costs related to such medical utilisation."
CRD COMMENTARY - Selection of comparators
The analysis compared MST with standard care within the authors' setting. Standard care was chosen as the comparator to represent current practice within the authors' setting and might differ in other settings.
Validity of estimate of measure of effectiveness
The authors used a randomised controlled trial to obtain their effectiveness data. Such a design helps to minimise systematic differences between patients in the two groups, thus improving the internal validity of the study. This was demonstrated through the demographic comparisons between the groups at baseline, which indicated that there were no statistically significant differences. The method of randomisation was not reported in the current study. However, the length of the study, loss to follow-up and blinding of the assessment were all reported, suggesting that the internal validity of the study is likely to be high. The analysis was conducted on an intention to treat basis and extensive statistical analyses took potential biases and confounding factors into consideration. However, no power calculations, to ascertain whether the results obtained were due to the intervention or to chance, were reported. The study sample was representative of the study population as it included adolescents with poor metabolic control. The authors used hospital admissions as a proxy for health outcomes, but acknowledged that this was not ideal.
Validity of estimate of measure of benefit
The authors did not estimate a summary measure of health benefit. The study is therefore categorised as a cost-consequences analysis.
Validity of estimate of costs
The costing was carried out from the hospital and third-party payer perspectives. The authors acknowledged that a comprehensive cost analysis was not conducted and the most important missing element was the cost of MST. Although MST reduced inpatient admissions, the resulting cost-savings may not have been sufficient to outweigh the additional costs associated with running the programme. The authors noted the need for longer run studies to estimate the full costs of MST. Nevertheless, the analysis conducted was clearly reported with unit costs and quantities given separately and broken down into their constituent parts. The use of charges to proxy costs has the limitation of not reflecting true opportunity costs, thus restricting the external validity of the results. Discounting was not relevant as the costs were incurred during less than 2 years. A price year was not reported, which will hamper any possible inflation exercises.
The authors could only make very limited comparisons of their work as there were few published studies exploring the ability to prevent recurrent diabetic ketoacidosis, which was one of the primary reasons for inpatient admissions. There were no comparisons of the cost-savings. The issue of generalisability was addressed with respect to the institutional perspective adopted and how costs may differ if the setting was altered. The results were presented in full and the authors provided a valuable discussion about the key cost-drivers that contributed to the overall cost-savings. The conclusions drawn were an accurate reflection of the results presented and related well to the scope of the study. Several limitations were noted. For example, the lack of generalisability and the fact that some patients may have attended a different hospital for emergency diabetic assistance and such attendance may not have been recorded for the purposes of this study.
Implications of the study
The authors did not make any recommendations for policy or practice but noted the need for longer run studies into the cost and effectiveness of MST.
Source of funding
Funded by the National Institute of Diabetes, Digestive and Kidney Diseases.
Ellis D A, Naar-King S, Frey M, Templin T, Rowland M, Cakan N. Multisystemic treatment of poorly controlled type 1 diabetes: effects on medical resource utilization. Journal of Pediatric Psychology 2005; 30(8): 656-666
Other publications of related interest
Ellis DA, Naar-King S, Frey MA, et al. Use of multisystemic therapy to improve regimen adherence among adolescents with type 1 diabetes in poor metabolic control: a pilot study. J Clin Psychol Med Settings 2004;11:315-24.
Ellis DA, Naar-King S, Frey MA, et al. Case study: feasibility of multisystemic therapy as a treatment for urban adolescents with poorly controlled type 1 diabetes. J Pediatr Psychol 2003;28:287-94.
Subject indexing assigned by NLM
Adolescent; Adult; Analysis of Variance; Diabetes Mellitus, Type 1 /economics /psychology /therapy; Diabetic Ketoacidosis /economics /prevention & Female; Health Care Costs; Health Services /utilization; Hemoglobin A, Glycosylated /metabolism; Humans; Male; Patient Compliance /psychology; Psychotherapy /methods; Self Care /psychology; control
Date bibliographic record published
Date abstract record published