The study examined three treatments for adults and children suffering from allergic rhinitis and asthma. The treatments were current symptomatic treatment (CST), injectable specific immunotherapy (SIT) and sublingual immunotherapy (SLIT). Both injectable SIT and SLIT were given in combination with CST. Injectable SIT and SLIT were given for 3 years in the adult population and for 4 years in the juvenile population.

Treatment.

The objective of the study was to assess the cost-effectiveness of the alternative strategies for treating allergic rhinitis and asthma in two populations (children and adults) with two different causes of allergy (dust-mite and pollen). Injectable SIT has been shown to be effective in reducing cases of rhinitis or asthma, but it is often associated with side-effects that could be reduced by sublingual administration (SLIT). A decision model defined by an expert panel was created to evaluate the costs and benefits of the three treatments using published evidence. The analysis was carried out from the perspective of the French health insurance system.

Cost-effectiveness analysis.

The study population comprised two cohorts of patients. One cohort was children over 5 years of age presenting with allergic rhinitis, with or without allergic asthma (juvenile model). The other cohort was young adults over 16 years of age presenting with allergic rhinitis, with or without allergic asthma (adult model).

A decision tree was constructed to simulate the natural history of the disease under each of the three treatments in both populations under examination (juvenile and adult models). The time horizon was 6 years for the adult population and 7 years for the juvenile population. Allergy could be due to either dust-mites (dust-mite model) or pollen (pollen model). A graphical representation of the model was provided, together with the main assumptions made. Different pathways were considered according to expert opinion.

The setting was secondary care. The economic study was carried out in France.

The dates to which the resource use and effectiveness data referred were not reported. The costs were derived from 2002 and 2003 databases.

The clinical data used in the model were:

the proportions of individuals with rhinitis or allergic asthma in the four models (juvenile, adult, dust-mite allergy and pollen allergy),

the distribution of severity levels, and

treatment efficacy (the numbers of improved patients and asthma cases).

Most of the clinical and epidemiological data were based on expert opinion. Some published studies and unpublished reports also provided data, although details of these were not given. In particular, the main source of treatment effect estimates was not clear.

Expert opinion was elicited using a Delphi panel of 11 members (10 allergologists and one epidemiologist). Information on the approach used to identify and select the published studies was not given.

The summary benefit measures used were the numbers of improved patients and asthma cases. Both measures were obtained using the decision model, on the basis of the pathways followed by patients. No discounting was required.

The cost analysis was carried out from the viewpoint of the French health insurance system. It included the direct medical costs associated with drugs, visits and diagnostic tests. The unit costs and the quantities of resources used were not presented separately. The drug costs came from a published database. Doctors' visits and tests were estimated using tariffs from the French Nomenclature Generale des Actes des Praticiens (NGAP). Resource quantities were based on expert opinion (Delphi panel) and recommendations from both French and international guidelines. The costs were discounted at an annual rate of 3% as long-term costs were evaluated. The price year appears to have been 2003, although this was not explicitly stated.

Productivity costs associated with asthma were considered only in the adult model, the number of workdays lost being derived from a published study. The productivity costs were valued using 2003 average daily allowances for sick-leave, as paid by the French Social Security. The unit costs and the resource quantities were not presented separately. As in the analysis of the direct costs, an annual discount rate of 3% was applied and 2003 costs were used.

Euros (EUR).

No statistical analyses of the costs were carried out.

A univariate sensitivity analysis was carried out to assess the robustness of the model results to variations in economic and epidemiological data. In particular, the official general practitioner's tariff for SIT injection from the NGAP nomenclature was introduced instead of the general practitioner's tariff used as current remuneration by allergologists. Alternative distributions of severity levels were derived from two published studies. Ranges of values defined by the Delphi panel were also used for other clinical data.

In a hypothetical cohort of 1,000 patients, the numbers of improved patients with CST, injectable SIT and SLIT, respectively, were:

272, 608 and 569 in the juvenile dust-mite allergy model;

158, 689 and 636 in the juvenile pollen allergy model;

230, 570 and 493 in the adult dust-mite allergy model; and

242, 662 and 592 in the adult pollen allergy model.

In a hypothetical cohort of 1,000 patients, the numbers of asthma cases with CST, injectable SIT and SLIT, respectively, were:

564, 362 and 385 in the juvenile dust-mite allergy model;

559, 185 and 193 in the juvenile pollen allergy model;

499, 306 and 341 in the adult dust-mite allergy model; and

420, 191 and 218 in the adult pollen allergy model.

The total direct costs with CST, injectable SIT and SLIT, respectively, were:

EUR 1,676, EUR 1,793 and EUR 2,381 in the juvenile dust-mite allergy model;

EUR 842, EUR, 1,065 and EUR 1,143 in the juvenile pollen allergy model;

EUR 1,285, EUR 1,361 and EUR 1,783 in the adult dust-mite allergy model; and

EUR 642, EUR 945 and EUR 986 in the adult pollen allergy model.

The indirect costs with CST, injectable SIT and SLIT, respectively, were:

EUR 1,476, EUR 800 and EUR 894 in the adult dust-mite allergy model; and

EUR 1,128, EUR 495 and EUR 560 in the adult pollen allergy model.

Incremental cost-effectiveness ratios were calculated in order to the combine costs and benefits of the alternative strategies.

The incremental cost per additional improved patient with injectable SIT over CST was EUR 349 for juvenile dust-mite allergy, EUR 420 for juvenile pollen allergy, EUR 224 for adult dust-mite allergy and EUR 722 for adult pollen allergy.

The incremental cost per additional improved patient with SLIT over CST was EUR 2,371 for juvenile dust-mite allergy, EUR 630 for juvenile pollen allergy, EUR 1,899 for adult dust-mite allergy and EUR 892 for adult pollen allergy.

The incremental cost per asthma case avoided with injectable SIT over CST was EUR 583 for juvenile dust-mite allergy, EUR 597 for juvenile pollen allergy, EUR 393 for adult dust-mite allergy and EUR 1,327 for adult pollen allergy.

The incremental cost per asthma case avoided with SLIT over CST was EUR 3,938 for juvenile dust-mite allergy, EUR 824 for juvenile pollen allergy, EUR 3,158 for adult dust-mite allergy and EUR 1,708 for adult pollen allergy.

The sensitivity analysis showed that the base-case results were robust to the variations considered in the sensitivity analysis, except when the official NGAP tariff for SIT injection was used. Under this scenario, injectable SIT became the dominant strategy compared with CST in dust-mite and pollen allergy.

Injectable specific immunotherapy (SIT) was a cost-effective treatment in pollen and dust-mite allergic rhinitis and asthma in comparison with current symptomatic treatment (CST). Sublingual SIT was cost-effective in pollen-induced rhinitis, especially in children.

The choice of the comparators appears to have been appropriate. The reasons for comparing SIT with SLIT (i.e. the different tolerability profile) were described. You should decide whether they are valid comparators in your own setting.

The use of modelling was appropriate in order to simulate the natural  course of the disease and its treatment.  The authors justified the  choice of a decision tree and its preference with respect to a Markov  model.  Given the differences in the population (adult versus juvenile)  and allergy (dust-mite versus pollen), four separate models were  considered.  However, the model structure was the same for all the four  cases and was depicted graphically.  The key assumptions of the model  were described and justified.

The effectiveness data used in the model were mainly derived from expert opinion, based on a Delphi process. Details of the process were not reported in detail, but it usually represents a rigorous approach to defining clinical inputs. The decision to use expert opinion was justified by the heterogeneity of estimates found in published studies. Some data were also derived from the literature, although details of a systematic search for data were not provided and there was no information on the primary studies. However, the authors stated that the data used in the model reflected the evidence reported in clinical trials.

The summary benefit measures were specific and not generalisable to other health care interventions. However, they are commonly used for patients with asthma. Both measures were derived using experts- opinions.

The analysis of the costs was consistent with the perspective of the analysis. All the relevant categories of costs were included. There was little information on the unit costs and quantities of resources used, which limits the possibility of replicating the analysis in other settings. The sources of the data were reported for all items. The costs were treated deterministically but the impact of using alternative cost estimates was investigated in the sensitivity analysis. The price year was implicitly reported, which will facilitate reflation exercises in other time periods.

The authors compared their findings with those from a German model and pointed out methodological differences between them. Comparisons with the results from other studies were also made. The issue of the generalisability of the study results to other settings was not explicitly addressed and few sensitivity analyses were carried out. Thus, the results of the study should be considered country-specific. The authors noted that the model did not consider treatment compliance, which might represent a key issue of the analysis. In general, the main limitations of the study appear to have been the weak analysis of uncertainty that would have been improved had a more sophisticated approach (such as a probabilistic analysis) been used, and the use of expert opinion to populate the model.

The study results support the use of SIT for the treatment of allergic rhinitis and asthma. Further epidemiological studies should be undertaken to corroborate the current findings.

None stated.

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.

Berto P, Passalaqua G, Crimi N, et al. Economic evaluation of sublingual immunotherapy vs symptomatic treatment in adults with pollen-induced respiratory allergy: the Sublingual Immunotherapy Pollen Allergy Italy (SPAIN) study. Ann Allergy Immunol 2006;97:615-21.

Schadlich PK, Brecht JG. Economic evaluation of specific immunotherapy versus symptomatic treatment of allergic rhinitis in Germany. Pharmacoeconomics 2000;17:37-52.

This record was compiled by CRD commissioned reviewers according to a set of guidelines developed in collaboration with a group of leading health economists.