|
Evaluating early dementia with and without assessment of regional cerebral metabolism by PET: a comparison of predicted costs and benefits |
Silverman D H S, Gambhir S S, Huang H W C, Schwimmer J, Kim S, Small G W, Chodosh J, Czernin J, Phelps M E |
|
|
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 health intervention examined in the study was a new non-invasive neurobiologically based approach using molecular imaging with PET for the evaluation of Alzheimer's disease (AD), the most common cause of dementia in the geriatric population.
Economic study type Cost-effectiveness analysis.
Study population The study population comprised geriatric patients with cognitive impairment, which could potentially represent the early manifestations of AD.
Setting The setting of the study was not clearly reported but appears to have been secondary care. The economic study was carried out in the USA.
Dates to which data relate Data on effectiveness evidence and resource use were gathered from studies published between 1985 and 2001. The price year was not clearly reported, but it appears to have been 2001.
Source of effectiveness data The effectiveness evidence was derived from a review of the literature, augmented by estimates made by the authors.
Modelling A decision-tree model was constructed to represent the conventional and PET approaches. The model was populated with data obtained from the literature.
Outcomes assessed in the review The outcome measures used as inputs in the decision model were the following probabilities: needing special tests or outside consultation, obtaining neuropsychological testing, having multiple cognitive domains affected at initial evaluation and at follow-ups, potentially reversible cause of dementia evident on history and physical examination or common laboratory tests, and identifying potentially reversible dementia with history and physical examination or with CT or MRI and completely reversing it with therapy. Other outcomes assessed were sensitivity and specificity of both PET and the conventional approach, both compared with neuropathologic confirmation.
Study designs and other criteria for inclusion in the review The study designs were reported for only three studies, which were defined as class I studies, indicating "a well designed prospective study in a broad spectrum of persons with the suspected condition, using a 'gold standard' for case definition, and enabling the assessment of appropriate tests of diagnostic accuracy". In general, studies to be included required at least some patients who were geriatric (older than 60 years) and with dementia (Mini-Mental State Examination scores of +/-20). Studies were excluded if they involved diagnostic work-ups lacking the majority of types of examination used in the context of either the conventional or the PET approach.
Sources searched to identify primary studies The MEDLINE and BIOSIS databases were searched for articles published in English between 1985 and 2000. The keyword "dementia" was cross-referenced with one or more of the following relevant terms: early diagnosis, early evaluation, Alzheimer or Alzheimer's, cognitive impairment, cognitive dysfunction, practice parameter, laboratory, delirium, depression, neuropsychological testing, CT, MRI, PET, sensitivity, specificity, follow-up, and reversible.
Criteria used to ensure the validity of primary studies Methods used to judge relevance and validity, and for extracting data Number of primary studies included Fourteen primary studies were used as sources of effectiveness evidence.
Methods of combining primary studies The method of combination was not reported. However, the authors reported that a conservative bias was permitted in favour of the conventional approach, when possible.
Investigation of differences between primary studies Results of the review The probability values were as follows:
0.035 for needing special tests or outside consultation,
0.59 for obtaining neuropsychological testing,
0.67 for having multiple cognitive domains affected at initial evaluation and 0.01 at follow-up,
0.22 for potentially reversible cause of dementia evident on history and physical examination or common laboratory tests,
0.045 for identifying potentially reversible dementia with history and physical examination and completely reversing it with therapy, and
0 for identifying potentially reversible dementia with CT or MRI, not found by physical examination or other common laboratory tests and then completely reversing it with therapy.
For the PET approach, sensitivity ranged from 88% to 95% and specificity from 67% to 73%.
For the conventional approach, sensitivity ranged from 49% to 96% and specificity from 50% to 100%.
Methods used to derive estimates of effectiveness Authors' experience was used to derive some effectiveness estimates where published data were not available or to make assumptions in the decision model.
Estimates of effectiveness and key assumptions The assessed probability of initial comprehensive history and physical examination was set as 1. The probability of needing CT or MRI was 0.625 (1 at initial evaluation and 0.10 in comprehensive follow-up evaluation).
Measure of benefits used in the economic analysis The benefit measures obtained from the decision model were the rates of false-positive and false negative and the overall accuracy. Only the overall accuracy rate was used in the economic analysis as the denominator of the cost-effectiveness ratio.
Direct costs No discounting was carried out, as the time horizon of the analysis was only 6 months. The cost/resource boundary adopted was that of the payer (Medicare). Quantities of resources used were not reported. The health service costs included in the cost analysis were history and physical examination (at initial stage and at follow-ups), MRI (with and without contrast, and neuropsychologic testing), laboratory tests (complete blood count, erythrocyte sedimentation rate, thyroxine, thyroid-stimulating hormone, basic metabolic panel, hepatic panel, folate, vitamin B12 level, syphilis serology, and brain PET), extra care needed for AD patients who progress past early stage while going untreated because of false-negative diagnosis, and one-year supply of cholinesterase inhibitor (such as donepezil and rivastigmine) unnecessarily prescribed to non-AD patients because of false-positive diagnosis. Reimbursement rates for each cost item were reported. The estimation of costs was generally based on 2001 Medicare outpatient reimbursement rates, including professional and technical components, when possible. The Medicare reimbursement rate for brain PET was obtained by multiplying the Medicare whole-body reimbursement rate by a factor of 0.70 in order to reflect the actual ratio of reimbursement rates from private insurance companies. The estimation of quantities of resources used was derived from the literature. The price year was not clearly reported, but appears to have been 2001.
Statistical analysis of costs No statistical analyses of costs were carried out.
Indirect Costs Indirect costs were not included.
Sensitivity analysis Sensitivity analyses were conducted to take into account the variability of data in different settings. The model inputs varied were cost, sensitivity, and specificity of PET, and cost of AD care. The type of analysis was not reported. A further sensitivity analysis was conducted varying model assumptions regarding the protocol of the conventional approach as recently suggested by the American Academy of Neurology (AAN) (MRI without contrast was obtained for 100% of patients evaluated and the price of syphilis serology testing was excluded from the items in the laboratory panel).
Estimated benefits used in the economic analysis The estimated false-positive and false-negative rates were 23.01% and 8.25% in the conventional approach and 12.04% and 3.14% in the PET approach. The overall accuracy rate was 68.74 in the conventional approach and 84.82 in the PET approach.
Cost results Mean cost per patient for evaluation and management was $3,564 for the conventional approach and $3,433 for the PET approach, with a modest cost-saving associated with the PET approach ($131).
Synthesis of costs and benefits Costs and benefits were combined by calculating the average cost per accurate diagnosis with the two approaches. No incremental analysis was performed. The cost per accurate diagnosis was $5,185 with the conventional approach and $4,047 with the PET approach. The cost-savings associated with the PET approach were $1,138. If a fixed total expenditure was available (i.e., $100,000), the number of correct diagnoses was 19 with the conventional approach and 25 with the PET approach. Variations of the PET costs indicated that the choice of the approach became cost-neutral only when the cost of PET was set at $2,728, which was 64% higher than the value used in the base analysis. The cost-effectiveness of the PET approach was not affected by variations in the other mentioned variables.
Authors' conclusions The authors concluded that a diagnostic approach for dementia based on PET may be more effective than the conventional approach, without increasing the overall costs of patients' evaluation and management. The PET approach led to a greater number of patients being accurately diagnosed for the same level of financial expenditure as the conventional AAN approach. Furthermore, the cost of a dedicated brain PET scan is generally less than the costs of one year of pharmacotherapy for unnecessary treatment of a patient who was misdiagnosed, and less than one month of productivity losses for a patient who was diagnosed when the disease was already in progress.
CRD COMMENTARY - Selection of comparators The rationale for the selection of the comparator was clear. The conventional approach was selected since it represented the routine strategy for the diagnostic evaluation of patients with suspected AD. In addition, the conventional approach was recommended by the AAN. You, as a user of this database, should assess whether the conventional approach may represent the routine diagnostic procedure in your own setting.
Validity of estimate of measure of effectiveness The effectiveness analysis was based on what appeared to be a systematic review of the available literature concerning AD plus some estimates made by the authors. Inclusion/exclusion criteria for the primary studies and search methods were clearly reported. However, although several effectiveness estimates were combined, the authors reported few details regarding the choice of the best estimate in the range obtained from the literature. To take into account uncertainty and variability of some effectiveness estimates, sensitivity analyses were conducted using the reported range of values.
Validity of estimate of measure of benefit The benefit measure used in the economic analysis (overall accuracy rate) was obtained through the decision model. It appears to have been appropriate to assess the diagnostic yield of the two approaches. However, since the disease considered has a strong impact on quality of life of patients (and caregivers), it would have been interesting to have used a benefit measure reflecting preferences of patients (or caregivers).
Validity of estimate of costs It appears that all categories of costs relevant to the perspective adopted were included in the analysis. Unit costs were reported for all health services considered in the costing. The estimation of costs, however, appeared quite specific to the study setting (Medicare reimbursement system). Variability in the cost data was appropriately dealt with in the sensitivity analyses. The price year was not clearly reported which makes reflation exercises to other settings problematic.
Other issues The authors made some comparisons of their findings with those from other studies. Although not specifically addressed, the generalisability of the study to other settings was addressed through the performance of sensitivity analyses on several model inputs and cost data.
Implications of the study The main implication of the study is that the use of PET allows disease to be diagnosed at an early stage and patients to be managed appropriately. This consideration raises the question of the correct timing of the performance of PET: the authors suggested that PET should be carried out "as soon as it has been determined that the patient is an appropriate candidate for the evaluation of cerebrocortical dysfunction".
Source of funding Supported by funds from the US Department of Energy; Los Angeles Alzheimer's Association, Turken Family Foundation Award; and the National Institutes of Health/National Institute on Aging.
Bibliographic details Silverman D H S, Gambhir S S, Huang H W C, Schwimmer J, Kim S, Small G W, Chodosh J, Czernin J, Phelps M E. Evaluating early dementia with and without assessment of regional cerebral metabolism by PET: a comparison of predicted costs and benefits. Journal of Nuclear Medicine 2002; 43(2): 253-266 Indexing Status Subject indexing assigned by NLM MeSH Aged; Algorithms; Alzheimer Disease /economics /physiopathology /radionuclide imaging; Cerebrovascular Circulation; Cost Savings; Cost-Benefit Analysis; Humans; Sensitivity and Specificity; Tomography, Emission-Computed /economics AccessionNumber 22002000448 Date bibliographic record published 31/08/2002 Date abstract record published 31/08/2002 |
|
|
|