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.

A statewide free nicotine replacement therapy (NRT) programme and a statewide smoke-free workplace policy were examined. Only nicotine patches and gums were included in the NRT programme, other medications (such as bupropion) were excluded.

Treatment.

Cost-effectiveness analysis and cost-utility analysis.

The study population comprised the Minnesota population of smokers.

The setting was the community. The study was carried out in the USA.

The clinical data were derived from studies published between 1993 and 2004. The costs and resource use data came from 2000 to 2005. The price year appears to have been 2002.

The effectiveness evidence was derived from a synthesis of completed studies and authors' opinions.

The outcomes estimated from the literature were:

the size of Minnesota's smoking population;

the effectiveness of a free NRT programme within the Minnesota population, which included the proportion of smokers and quitters as well as relapse rates and quality-adjusted life-years (QALYs) associated with quitting; and

the effectiveness of a smoke-free workplace policy.

Quitting was defined as continued cessation after 6 months.

Much of the data were derived from published studies. However, limited information on the design and characteristics of the primary studies was provided. Data on the Minnesota population were derived from Census data. Other data came from a MPAAT survey and the National Health Interview survey. The effectiveness data for NRT were obtained from clinical trials. A meta-analysis was also used for the absolute smoking reduction produced by a smoke-free workplace policy.

Not stated.

Not stated.

Not stated.

Fourteen primary studies provided clinical evidence.

Each study provided primary estimates that were not pooled with data derived from other sources.

Not stated.

The total Minnesota population aged 18 years and older was 3,717,580 in 2001, with a smoking prevalence of 22.2%.

The proportion of Minnesotan smokers who were actively considering quitting was 64%.

Eighty per cent of the current smokers who were interested in quitting would use NRT if cost were no object.

The rate of former and current smokers that used NRT when they last attempted to quit was 15%.

The net gain in NRT users obtained by providing free NRT would be 68% (i.e. 80% multiplied by 85%) of potential quitters or 359,000 people.

It was estimated that 54.9% of Minnesotan smokers were moderate to heavy smokers (> 15 cigarettes/day), thus 197,000 of the 359,000 new NRT users could have gained benefit from NRT.

The pooled odds ratio of quitting with NRT over baseline was 1.71 (95% confidence interval, CI: 1.56 - 1.87).

The baseline quit rate for smokers interested in quitting was 10.6%.

The smoking relapse rate was 35%, resulting in a total of 18,500 quitters after one year of the free NRT programme.

Each sustained quitter generated 1.58 discounted (at 3%) QALYs for an average quitter aged 45 years.

The number of indoor workers in Minnesota was 2,166,000.

The rate of indoor workers in Minnesotan already covered by smoke-free workplace policies was 73.9%. Thus, there were 125,000 smokers who worked indoors and were not covered by a smoke-free workplace policy.

The implementation of smoke-free workplaces would produce an absolute smoking prevalence reduction of 3.7%. Thus, a smoke-free workplace policy would lead to a 16.7% decline in indoor workers who smoked.

The quit rate, regardless of a smoke-free workplace policy, was 2.5% per annum. Thus, a smoke-free workplace policy would lead to an additional 14.2% in the quit rate among indoor workers who smoked.

The rate of compliance with smoke-free workplace policies was 90%.

The total number of quitters generated was 10,400.

The authors made some assumptions to derive effectiveness estimates.

NRT was not effective in helping light smokers (less than 15 cigarettes/day) to quit. Two quit attempts within one year were considered. Relapsing individuals gained no QALY benefit. The gain in QALYs ceased after the age of 65 years, given that the benefit of smoking cessation in life expectancy is mainly due to the prevention of early heart disease.

The summary benefit measures used were the total expected quitters and the QALYs. The former measure (total expected quitters) was used in the cost-effectiveness analysis, while the latter (QALYs) was used in the cost-utility analysis. The QALYs were discounted at an annual rate of 3%. The authors stated that they converted the benefit of quitting into QALYs in accordance with the existing literature, but no other details were given.

The cost/resource boundary of the service payer was adopted. The cost of a free NRT programme was estimated by calculating the cost per quit attempt, which was based on the quantities of resources used derived from national data and average wholesale prices for nicotine gum and nicotine patches. The costs of administering a free NRT programme, such as advertising, coordination, dispensation, or counselling were not included. The unit costs were presented separately from the quantities of resources used for most items of the NRT programme. The costs of implementing statewide smoke-free workplaces were associated with enactment of the policy and its enforcement. The costs associated with enactment were derived from the costs of running a smoke-free workplace policy in Florida and included the costs of personnel, office expenses/collateral materials, legal services, public opinion research, paid petition gathering, and paid media. The costs associated with enforcement came from media campaign expenditures in Florida. The authors stated that both sources of costs tended to overestimate the costs of the two statewide programmes in Minnesota. The unit costs were not presented separately from the quantities of resources used since most costs were reported as macro-categories for the smoke-free workplace programme. The price year appears to have been 2002. Discounting was not relevant since one-year costs were estimated.

The costs were treated deterministically.

The indirect costs were not included in the economic evaluation.

US dollars ($).

The robustness of the base-case results of the cost-effectiveness and cost-utility analyses was tested under three alternative scenarios. In the first scenario, the free NRT programme was provided only to heavy smokers. In the second scenario, light smokers experienced full cessation benefits after receiving NRT without suffering adverse effects from excess nicotine. The third scenario assumed a low compliance (50%) with smoke-free workplace policies. A probabilistic sensitivity analysis was performed using a Monte Carlo simulation to estimate the distribution of the costs and benefits of both programmes (10,000 trials were generated). The authors set the ranges of values used.

The QALYs were estimated from the literature. Each sustained quitter aged 45 years generated 1.58 QALYs.

The total expected annual number of quitters in Minnesota in 2001 was 18,500 with the free NRT programme and 10,400 with a smoke-free workplace programme.

The total expected costs in Minnesota in 2001 were $130,00,000 with a free NRT programme and $8,300,000 with a smoke-free workplace programme. Thus, the free NRT programme was 15 times more expensive than the smoke-free workplace programme.

Average cost-effectiveness ratios (i.e. cost per quitter) and average cost-utility ratios (i.e. cost per QALY) were calculated to combine the costs and benefits of the two programmes.

The average cost per quitter was $7,020 with the free NRT programme and $799 with the smoke-free workplace programme.

The average cost per QALY was $4,440 with the free NRT programme and $506 with the smoke-free workplace programme.

The sensitivity analysis showed that the cost per quitter and the cost per QALY did not vary substantially when the clinical and economic parameters were varied. Clearly, when compliance with the programme dropped, higher cost-effectiveness and cost-utility ratios were obtained. If the free NRT programme was restricted to heavy smokers, or if all light smokers received the same smoking cessation benefit from NRT as heavy smokers did, the average cost per quitter was $3,790 and the average cost per QALY was $2,400.

The probabilistic sensitivity analysis confirmed the robustness of the results of the cost-utility and cost-effectiveness analyses. In particular, the Monte Carlo simulation showed overlap in the CIs of quitter estimates for both the free NHT programme and a smoke-free workplace policy, but no overlap was found in the cost or cost-effectiveness ratio estimates.

Both smoking cessation programmes were cost-effective. The free nicotine replacement therapy (NRT) programme led to more quitters, but its higher costs resulted in the smoke-free workplace programme being more cost-effective.

The rationale for the selection of the comparators was clear, as two widely used strategies to improve quit rates among smokers were used. You should decide whether they are valid comparators in your own setting.

The effectiveness evidence came from published studies, although it was not explicitly stated whether a systematic review of the literature was undertaken. The primary estimates appear to have been identified selectively. Details of the design of the primary studies and the patient samples. Thus, it was difficult to assess the validity of the primary sources of clinical evidence. However, data on the effectiveness (quitting rate) of the two programmes were obtained from clinical trials and a meta-analysis, thus enforcing the internal validity of the primary sources. Other data came from US and federal statistics. Differences between the primary studies were not investigated. Some assumptions were also made. The issue of uncertainty was addressed in the sensitivity analysis.

Both a specific and a general benefit measure were used in the economic evaluation. This enables comparisons with the benefits of other health care interventions to be made. The approach used to calculate the QALYs was reported, but the source of the utility values was not given.

The authors did not state explicitly which perspective was adopted in the study. It appears that the costs have been estimated from the perspective of the service payer. The unit costs were not presented separately from the quantities of resources used for all items since some costs were given as macro-categories. The source of the costs was provided and, owing to the lack of economic estimates from Minnesota, data from other US states were used. The authors stated that the costs could have been overestimated. The cost calculation was explicitly described. The price year was reported, which aids reflation exercises in other settings. The costs were treated deterministically in the base-case, but some categories of costs were assigned a probability distribution in the sensitivity analysis.

The authors stated that their findings were consistent with those from published studies. They did not investigate the generalisability of their results to other settings, but did carry out several sensitivity analyses. The issue of uncertainty that surrounded both the clinical and cost estimates was explicitly addressed in the probabilistic sensitivity analysis. The authors stated that the analysis was restricted to a one-year time horizon because of the lack of reliable data required to project the results of the analysis to a longer timeframe. It was also noted that the estimation of benefits did not account for the reduction in passive smoking exposure among non-smoking indoor workers, or the additional benefits that media campaigns have on non-indoor workers, such as the prevention of smoking initiation, and smoking cessation among other populations.

The study results suggested that smoke-free workplace campaigns should be a priority for public health programmes.

Supported by a National Cancer Institute grant and a Health Resources and Services Administration training grant.

Silagy C, Mant D, Fowler G. Lodge M, Meta-analysis on efficacy of nicotine replacement therapies in smoking cessation. Lancet 1994;343:139-42.

Fichtenberg CM, Glantz SA. Effect of smoke-free workplaces on smoking behaviour: systematic review. BMJ 2002;325:188.

Farrelly M, Evans W, Sfekas A. The impart of workplace smoking bans: results from a national survey. Tobacco Control 1999;8:272-7.