|
Cost-effectiveness of combined outreach for the pneumococcal and influenza vaccines |
Weaver M, Krieger J, Castorina J, Walls M, Ciske S |
|
|
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 investigated the implementation of a community-based outreach intervention to promote pneumococcal and influenza vaccines in elderly people.
Study population The study population comprised people aged 65 years or older who were residing in the targeted ZIP code areas and who returned a survey questionnaire. There were no exclusion criteria.
Setting The setting was community care. The economic study was carried out in the USA.
Dates to which data relate The effectiveness evidence was derived from literature published between 1994 and 2000. The costs were collected from a clinical trial published in 2000 that was conducted between September 1996 and March 1997. The price year was not reported.
Source of effectiveness data The effectiveness data for the outreach intervention were derived from a randomised controlled trial (RCT). Other data were derived from published studies of vaccine effectiveness.
Study sample The effectiveness data for the outreach intervention were derived from an RCT. Power calculations were not reported, although further details may be available in the parent study (Krieger et al., see Other Publications of Related Interest). The participants were selected from a senior centre and a marketing database that distinguished residents in five ZIP code areas. These people were sent an initial invitation letter and baseline questionnaire to determine their immunisation status. Those who returned the questionnaire and who met the inclusion criteria were randomised to the intervention or control group. A total of 1,246 people completed the baseline survey and 1,083 completed the follow-up survey. It was reported that the low-intensity intervention comprised 623 participants at baseline and 553 at follow-up, while the high-intensity intervention comprised 622 participants at baseline and 530 at follow-up.
Study design The study was a community-based RCT that was conducted in a single centre from September 1996 to March 1997. Randomisation to the intervention or control group was conducted by systematic allocation of alternate respondents to the survey. Blinding was not reported. The loss to follow-up was 70 in the intervention group and 92 in the control group. Reasons for the loss to follow-up were not given, but may have been reported in the parent study.
Analysis of effectiveness The primary health outcome used in the effectiveness analysis was self-reported vaccination rates (effectiveness of the intervention). As the parent trial found that vaccination rates varied by vaccine and prior immunisation status, two sets of estimates for the effectiveness of the intervention were calculated. One was for the intervention as implemented (full sample), while the other was for an intervention targeted at people who had never received the pneumococcal vaccine or those who had not received the influenza vaccine in the previous year (targeted sample).
It was unclear whether the analysis was conducted on an intention to treat basis or on treatment completers only. The present study did not state whether the groups were comparable at baseline. This information may have been reported in the parent study.
Effectiveness results The first analysis (full sample) showed that the intervention was 15% effective in promoting the pneumococcal vaccine (confidence interval, CI: 10 - 21) and 6% effective in promoting the influenza vaccine (CI: 2 - 11).
In the second analysis (targeted sample), the intervention was 21% effective in promoting the pneumococcal vaccine (CI: 14 - 28) and 27% effective in promoting the influenza vaccine (CI: 14 - 40).
Clinical conclusions The results showed that the outreach intervention became more effective in promoting the vaccines when targeted at those who had never received the pneumococcal vaccine, or those who had not received the influenza vaccine in the previous year.
Modelling A decision tree was developed for each independent cost-effectiveness analysis conducted. The analyses conducted were pneumococcal vaccine against pneumococcal bacteraemia, and influenza vaccine against influenza and pneumonia. Decision trees were developed to estimate the costs and benefits resulting from the adoption of each of the vaccine interventions. Although it was not clearly stated by the authors, it would appear that low- versus high-intensity interventions were compared in each cost-effectiveness analysis, with the low-intensity intervention being the control strategy (comparator). Therefore, the probability of getting a vaccine was higher for the intervention group than for the control group. The decision tree for bacteraemia had 6 annual cycles, whereas the decision tree for the influenza vaccine had 1 annual cycle. For the cost-effectiveness analysis of the combined outreach initiative, the authors calculated the sum of the costs and effectiveness of each independent analysis, except for the cost of the intervention, which was counted only once.
A number of assumptions were used in the construction of this model. First, it was assumed that the effectiveness of the intervention for the sub-sample analyses could be extended to a sample as large as the full sample in a targeted outreach initiative. Second, the vaccine affected the incidence of illnesses but not the severity when they occurred.
Outcomes assessed in the review Other health outcomes were derived from published studies. These related to the effectiveness of vaccines in preventing illness, the side effects of vaccines, the incidence of pneumococcal bacteraemia and influenza, and death rates.
Study designs and other criteria for inclusion in the review Estimates of effectiveness (with CIs) of pneumococcal vaccine were derived from a published study (Sisk et al., see Other Publications of Related Interest), which based estimates on a case-control study (Shapiro et al., see Other Publications of Related Interest). Estimates of effectiveness (with CIs) of influenza vaccine were derived from an RCT (Govaert et al., see Other Publications of Related Interest).
Sources searched to identify primary studies 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 Five primary studies were included in the review.
Methods of combining primary studies The results of the primary studies were not combined.
Investigation of differences between primary studies Results of the review The estimates of vaccine effectiveness for pneumococcal vaccination against bacteraemia (6-year cycle) were as follows:
year 1, 82% (CI: 56 - 93);
year 2, 76% (CI: 41 - 91);
year 3, 70% (CI: 26 - 89);
year 4, 65% (CI: 17 - 87);
year 5, 59% (CI: 10 - 85); and
year 6, 54% (CI: 3 - 83).
This was based on the weighted average for the 65 - 74 and 75 - 84 year age strata (weights were the number of people in each 10-year age interval in a stationary population).
The estimate of vaccine effectiveness for the influenza vaccine (1-year cycle) was 50% (CI: 39 - 65).
The incidence of bed-disability day with pneumococcal vaccine was 0.00005%.
The proportion of case of bacteraemia covered by vaccines serotypes was 88%.
The incidence of bacteraemia was 68.6/100,000.
The death rate from bacteraemia was 19.2/100,000.
The probability of dying when bacteraemia was eliminated was 4.4954%.
The incidence of a non-bed-disability day with influenza vaccination was 5%.
The probability of hospitalisation with influenza vaccination was 1.55%.
The probability of a bed-disability day from pneumonia and influenza was 13.84%.
The death rate from pneumonia and influenza was 126.7/100,000.
The probability of dying when pneumonia and influenza were eliminated was 4.3731%.
The death rate from all causes was 3,833.7/100,000.
The probability of death from all causes was 4.5173.
Measure of benefits used in the economic analysis The measure of benefit used was the quality-adjusted life-years (QALYs). There was no information on the methods used to evaluate the quality of life weights. A rate of 3% was used to discount the future benefits.
Direct costs The direct costs relating to the delivery of the community intervention were those for staff, volunteers, advisory committee, computer tracking system and materials. Research costs associated with the RCT were excluded. Only a few resource quantities and costs were reported separately. Staff, volunteer and advisory committee resource use was recorded by way of time sheets and other records. Resource use information on the computer tracking system and materials was not given. Staff time was valued at salary rates, plus benefits, and included a 7% mark up for general overheads (the rate for Public Health). This was calculated on the basis that 37.5% of all time at meetings was spent on the intervention. Volunteer rates were calculated on the basis of mean weekly earnings of people aged 65 years or older ($9.80 per hour). Advisory committee costs were also valued at $9.80 per hour, plus benefits where applicable. Again, 37.5% of time at meetings was the assumed input to the intervention. The computer tracking system and materials were valued at purchase price, with amortisation of the former calculated over 5 years.
The cost of the vaccines (derived from Public Health primary data), which were kept separate from the intervention costs, were $8.02 for the pneumococcal vaccine and $1.58 for the influenza vaccine. Participant expenses (derived from participant survey data) relating to receipt of the vaccine were $10.81 (pneumococcal) and $10.26 (influenza), respectively. The costs of side effects and consequent hospitalisation or outpatient visits relating to the two vaccines were recorded from published literature and were reported in more detail in the paper.
The costs were discounted at a rate of 3%. The price year was not reported.
Statistical analysis of costs The data were partially stochastic to address uncertainty (sensitivity analysis) in values for the effectiveness of the intervention and vaccines. Quasi-CIs were calculated for cost-effectiveness ratios, based on the CIs of the estimates of effectiveness. Other cost data were deterministic.
Indirect Costs There was no reported productivity or intangible cost.
Sensitivity analysis A one-way sensitivity analysis was carried out for the cost-effectiveness of the vaccines and outreach intervention using quasi-CIs calculated from the partial stochastic analysis. This analysis employed the upper and lower bounds detailed earlier.
Estimated benefits used in the economic analysis When considering the intervention as implemented, the total QALYs lost in the intervention and control groups were:
with pneumococcal vaccination, 1,040.62 for the intervention versus 1,041.08 for the control;
with influenza vaccination, 184.29 for the intervention versus 184.47 for the control; and
with the combined outreach initiative, 1,224.91 for the intervention versus 1,225.55 for the control.
When considering the targeted intervention, the total QALYs lost in the intervention and control groups were:
with pneumococcal vaccination, 1,041.05 for the intervention versus 1,041.70 for the control;
with influenza vaccination, 185.41 for the intervention versus 186.23 for the control; and
with the combined outreach initiative, 1,226.46 for the intervention versus 1,227.93 for the control.
The marginal effectiveness was calculated, showing the difference in QALYs between the intervention and control groups.
The marginal effectiveness was 0.46 for the pneumococcal vaccine, 0.18 for the influenza vaccine, and 0.64 for the outreach intervention when considering the intervention as implemented.
The marginal effectiveness was 0.65 for the pneumococcal vaccine, 0.82 for the influenza vaccine, and 1.47 for the outreach intervention when considering the targeted intervention.
The benefits were calculated over a 1-year period for the influenza vaccine and over a 6-year period for the pneumococcal vaccine.
Cost results The total intervention cost was $25,076.
When considering the intervention as implemented, the total costs in the intervention and control groups were as follows:
with pneumococcal vaccination, $43,627 for the intervention versus $18,903 for the control;
with influenza vaccination, $64,989 for the intervention versus $41,395 for the control; and
with the combined outreach initiative, $81,510 for the intervention versus $58,730 for the control.
When considering the targeted intervention, the total costs in the intervention and control groups were as follows:
with pneumococcal vaccination, $43,956 for the intervention versus $19,373 for the control;
with influenza vaccination, $74,376 for the intervention versus $55,971 for the control; and
with the combined outreach initiative, $91,657 for the intervention versus $74,390 for the control.
The marginal cost was roughly the same in all analyses.
The costs of side effects were taken into consideration.
The authors also projected that the cost of the intervention could be reduced to $19,315 per year by rearranging the project coordinator's input over a shorter period of time, and by switching to a manual data recording system.
Synthesis of costs and benefits Six estimates were presented for incremental costs per QALY saved. The first three related to cost-utility ratios for the pneumococcal vaccine, influenza vaccine and combined outreach initiative for the intervention as implemented (full sample). The incremental cost per QALY saved was $53,547 with the pneumococcal vaccine, $130,908 with the influenza vaccine, and $35,486 with the combined outreach initiative.
The second three cost-utility ratios were for the targeted sample. The incremental cost per QALY saved was $38,030 with the pneumococcal vaccine, $22,431 with the influenza vaccine, and $11,771 with the combined outreach initiative.
The results of the sensitivity analysis showed that the combined outreach intervention remained cost-effective, on the basis of a threshold of $47,590 and other thresholds proposed in published literature. From the literature, a cost-effectiveness ratio of less than $28,000 per QALY saved is classed as strong evidence, $28,000 - $140,000 per QALY saved is classed as moderate evidence, and more than $140,000 per QALY saved is classed as weak evidence.
Authors' conclusions The combined outreach intervention was reasonably cost-effective. The cost utility ratio was reduced by lowering the costs of the intervention, or by targeting the intervention to those who had never received a pneumococcal vaccine or to those who had not received the influenza vaccine within the last year.
CRD COMMENTARY - Selection of comparators It was not clear why the comparator(s) used were chosen, because the authors did not provide a justification for their choice. Indeed, it was also unclear what differences existed in content between the low- and high- intensity community outreach interventions proposed in the study schedule. You should decide if these represent widely used interventions in your own setting.
Validity of estimate of measure of effectiveness The analysis was based on a community-based RCT, which was appropriate given the study question. The choice of patient sample (those over 65 years old) was justified on the basis of current public health recommendations. It was unclear whether the study sample was representative of the study population, as those recruited were survey responders. No analysis of nonresponders was reported. The internal validity of the study was limited, given the method of randomisation, the lack of information on the comparability of the groups, and the failure to report reasons for loss to follow-up. Although an RCT was conducted, the analysis appears to have been based on those who remained at follow-up. In addition, no account was taken of potential confounding factors that inevitably present as part of many community-based interventions.
Some input parameters were also taken from a review of primary studies. The authors did not state whether or not they undertook a systematic review of the literature. Since the methods used to find and select the primary studies and to extract the data were unclear, it was difficult to assess the validity of the estimates. The authors appear to have used data from the available studies selectively. These facts potentially limit the reliability of the findings. However, most of the model parameters were varied in a sensitivity analysis, and this enhances the validity of the results.
Validity of estimate of measure of benefit The measure of benefit was health utility (QALYs). Failure to report how this was measured and valued contributes further uncertainty to the validity of the study findings. For the purposes of this abstract, the type of study was a cost-utility analysis (and not a cost-effectiveness analysis as the authors reported).
Validity of estimate of costs The authors claimed that they performed the cost analysis from a societal perspective, although the indirect costs were not included. For the purposes of this abstract, the analysis was actually conducted from the perspectives of the health care system and patient. In this respect, it appears that relevant categories of costs have been included. It was not possible to separate all the costs and the quantities, therefore reworking in another setting would be only partially achievable. The resource quantities were obtained largely from self-completed time sheets and other records which are subject to uncertainty. The unit costs were taken from a mixture of published sources, research settings and market value. A sensitivity analysis was, however, conducted to show the effect on the future intervention costs of different patterns of personnel input and computer-aided support. Discounting was appropriately conducted at 3% in the base-case analysis and 5% in the sensitivity analysis, but the failure to report a price year will limit any future reflation exercises.
Other issues As this was the first economic analysis of a community-based outreach intervention in this topic area, the authors did not compare their findings with other studies. The generalisability to the study population and other settings was unclear. The results of the analysis were adequately reported and the authors' conclusions reflected the scope of the analysis. The authors acknowledged several limitations of their study. First, the omission of unrelated medical costs of survivors from the sensitivity analysis. Second, the reliance upon community-specific studies for data on the incidence of and mortality from pneumococcal bacteraemia.
Implications of the study There were no specific recommendations for future practice or research.
Source of funding Funded by the US Centers for Disease Control and Prevention, Atlanta (GA), cooperative agreement U50/CCU011820-02 (Urban Research Centers), and United Way of King County, Seattle (WA).
Bibliographic details Weaver M, Krieger J, Castorina J, Walls M, Ciske S. Cost-effectiveness of combined outreach for the pneumococcal and influenza vaccines. Archives of Internal Medicine 2001; 161: 111-120 Other publications of related interest Krieger J, Castorina J, Ciske S, et al. Increasing influenza and pneumococcal immunization rates: a randomised controlled study of a senior center-based intervention. American Journal of Preventive Medicine 2000;18:123-31.
Sisk JE, Moskowitz AJ, Whang W, et al. Cost-effectiveness of vaccination against pneumococcal bacteremia among elderly people. JAMA 1997;278:1333-9.
Govaert TM, Thijs CT, Masurel N, et al. The efficacy of the influenza vaccination in elderly individuals: a randomized double-blind placebo-controlled trial. JAMA 1994;272:1661-5.
Shapiro ED, Berg AT, Austrian R, et al. The protective efficacy of polyvalent pneumococcal polysaccharide vaccine. New England Journal of Medicine 1991;325:1453-60.
Indexing Status Subject indexing assigned by NLM MeSH Aged; Cost-Benefit Analysis; Decision Trees; Humans; Influenza Vaccines; Influenza, Human /prevention & Models, Economic; Pneumococcal Infections /prevention & Pneumococcal Vaccines; Preventive Health Services /economics /organization & Quality-Adjusted Life Years; Washington; administration; control; control AccessionNumber 22001008019 Date bibliographic record published 31/08/2005 Date abstract record published 31/08/2005 |
|
|
|