Utilization and adherence to smoking cessation pharmacotherapy: A population-based study in Spain

Abstract

Objective

To analyse the utilisation of publicly funded smoking cessation pharmacotherapy within a public health system and to determine the proportion of quit attempts completing the recommended treatment.

Design

Retrospective population-based study.

Site

Public health system of Castilla y León (Spain), covering more than two million inhabitants, 2020–2024.

Participants

Individuals of any age who received reimbursed varenicline, bupropion or cytisinicline for smoking cessation during the study period.

Interventions

None; observational study of reimbursed pharmacotherapy.

Main measurements

Number of quit attempts, pharmacotherapy type, packs dispensed per attempt, age, sex, income-based co-payment status and area of residence. Adherence was indirectly estimated according to completion of the recommended treatment course.

Results

A total of 57,820 individuals used pharmacotherapy for smoking cessation, representing 14.5% of the estimated smoking population, and undertook 67,255 quit attempts. Overall utilisation was limited and adherence to longer treatment regimens was suboptimal. Treatment completion occurred in 26.8% of varenicline and 32.4% of bupropion courses, while cytisinicline predominated. Women made more repeated quit attempts and showed slightly better adherence than men (p < 0.001). Pharmacotherapy use was proportionally higher in rural areas, although lower among women in this setting. Lower consumption was observed during summer months and December, without statistically significant differences.

Conclusions

In routine clinical practice, the use of smoking cessation pharmacotherapy remains limited, with low completion of recommended treatment regimens. Differences by sex and socioeconomic status, together with suboptimal treatment completion, highlight the need to strengthen follow-up strategies and improve equity and accessibility within publicly funded cessation programmes.

Introduction

Tobacco smoking is a chronic, relapsing addictive disease that encompasses physical and psychological dependence, learned behaviour and social dependence.1, 2

Nicotine is the primary substance responsible for the development and maintenance of tobacco addiction. Repeated and prolonged exposure to this highly addictive and toxic compound leads to the development of tolerance.

Consequently, during the early stages of a quit attempt, individuals experience a withdrawal syndrome characterised by intense craving, anxiety, irritability, restlessness and impaired concentration.²

The combination of cognitive–behavioural therapy and environmental modification³ with treatment for nicotine dependence has been shown to increase the likelihood of successful smoking cessation.4, 5

In this context, the use of nicotine replacement therapy, varenicline, bupropion or cytisinicline is recommended, all of which are effective and safe options.⁶ However, only the latter three require medical prescription.⁷

Public reimbursement of smoking cessation treatments increases the number of individuals who successfully quit smoking,⁸ reduces socioeconomic barriers⁹ and improves equity in access to care.⁶

Following their inclusion in the health service, it is important to evaluate the reach and implementation of this measure within tobacco control strategies.7, 10

The aim of this study was to analyse the utilisation of funded smoking cessation pharmacotherapy and to identify the proportion of quit attempts that completed the recommended treatment.

Materials and methods

Study design and population

The present study builds on the methodology of a previous investigation¹¹ and extends it to the entire public health system in Castilla y León (Spain). A retrospective analysis was conducted on prescriptions for smoking cessation pharmacotherapy issued in a regional public health system between 2020, when these treatments were first publicly reimbursed, and 2024, inclusive. In 2024, the covered population comprised 2,388,350 inhabitants, with minimal variation between 2020 and 2025 (−0.04%).12, 13

The study cohort comprised 2,321,475 individuals holding public health insurance card,¹⁴ representing approximately 97% population coverage. Of these, 1,329,378 resided in urban areas, 221,945 in semi-urban areas and 770,152 in rural areas.¹³

According to the National Health Survey, an estimated 399,000 people aged 15 years or older were current smokers (240,100 men and 158,800 women).¹⁵

Dispensing data for varenicline, bupropion and cytisinicline were obtained from CONCYLIA, the regional pharmacy information system of Castilla y León, which provides anonymised individual-level data through unique identification codes.

Eligibility criteria

Individuals of any age were included if they had made at least one smoking cessation attempt in the previous year, reported a daily consumption of at least 10 cigarettes, scored ≥7 on the Fagerström Test for Nicotine Dependence, and were in the preparation or action stage of change. Pharmacological treatment was prescribed through the electronic prescription system.¹⁶ The clinical inclusion criteria form part of the requirements established for the prescription of these treatments within the public health system.⁷ In the electronic health record system, these criteria must be recorded and validated in accordance with national regulatory requirements¹⁷ so that the prescription module can issue electronic prescriptions. However, this clinical information is not directly available in the dispensing database used (CONCYLIA), and therefore, could not be verified individually in this study.

Individuals without public health insurance coverage or those who attempted to quit smoking without reimbursed pharmacotherapy were excluded. The prescribing module does not allow these medicines to be prescribed unless the eligibility criteria for reimbursement are met and recorded in the electronic health record.¹⁶

Variables

Data collected included the date of prescription and dispensing, the prescribed medication, and the number of packs dispensed per quit attempt. In addition, the number of quit attempts per individual, age, sex (binary variable: male/female), co-payment scheme according to income level based on the individual health card classification (TSI categories 1–5), as well as the area of residence, were recorded.

Statistical analysis

Analysis was performed using SPSS© version 29. The normality of continuous variables was assessed using the Kolmogorov–Smirnov test. The Mann–Whitney U test was used to analyse sex-related differences in age, number of quit attempts and monthly distribution of dispensed packs. The Fisher's exact test was applied to assess the association between sex and area of residence, treatment adherence according to the number of packs dispensed per pharmacological agent, and income level based on the TSI code; in the latter case, the Monte Carlo method was additionally applied.

All analyses were conducted using 95% confidence intervals, and a p-value <0.05 was considered statistically significant.

Ethical considerations

The study was approved by the Research Ethics Committee with Medicines of the Zamora Healthcare Area in 2024 (registration number 657). Given the retrospective design and the use of previously anonymised data, the requirement for informed consent was waived.

Results

During the study period, 57,863 individuals used smoking cessation pharmacotherapy. Of these, 57,820 individuals (47.7% women) were included in the analysis, representing 14.5% of the estimated smoking population in the region.¹⁵

These individuals undertook a total of 67,255 smoking cessation attempts. A single attempt was made by 50,105 individuals, while 7715 carried out two or more attempts (Fig. 1). Overall, 85,159 medication packs were dispensed, including 8298 of bupropion, 48,097 of cytisinicline and 28,764 of varenicline. The number of packs dispensed per quit attempt according to pharmacological agent is shown in Table 1.

Figure 1.

Study flow diagram. Study flow diagram showing inclusion and exclusion of participants and the number of quit attempts during the study period. A total of 43 individuals who used different pharmacological agents within the same smoking cessation attempt were excluded from the analysis. The study included 57,820 individuals who made smoking cessation attempts using prescription-only pharmacological treatments for nicotine dependence between 2020 and 2024. Of these, 50,105 individuals made a single quit attempt, while the remainder made two or more attempts. Overall, 67,255 smoking cessation attempts were recorded during the study period.

Table 1.

Number of packs dispensed per quit attempt by pharmacological agent.

Number of packs dispensed per quit attempt (n = 85,159)	Bupropiona	Cytisiniclineb	Vareniclinec
1	2299	47218	6374
2	1506	413	4038
3	532	14	4010
4	225	1	450
5	45	–	59
6	31	–	22
≥7	10	1	7

Overall, a total of 67,255 quit attempts were recorded during the study period, with 85,159 packs of pharmacological treatments dispensed.

^aBupropion: prescribed for smoking cessation; availability was limited from 2022 until the second half of 2024.

^bCytisinicline: publicly funded from February 2023 onwards.

^cVarenicline: withdrawn from the market in the first half of 2021 and reintroduced in the second half of 2024.

Varenicline was used in 14,960 quit attempts, of which 26.8% completed the recommended treatment (three packs). A total of 69.6% collected fewer than three packs, and 3.6% extended treatment beyond the recommended duration. Bupropion was used in 4648 attempts; 32.4% completed the recommended treatment (two packs), 49.5% collected a single pack, and 18.1% prolonged treatment. Cytisinicline was used in 47,647 attempts, of which only 0.9% extended treatment beyond one pack.

Cytisinicline was the most frequently used pharmacological agent in successive quit attempts (4857 in the second attempt, 916 in the third, 179 in the fourth, 29 in the fifth and 6 in the sixth).

The median age was 51 years, and the mean age was 50.5 years (standard deviation [SD] 11.8), with significantly higher values observed among women (p < 0.001) (Table 2). Twenty-three individuals were younger than 18 years, while 5729 were aged 66 years or older, including 106 aged 80 years or above.

Table 2.

Descriptive and analytical characteristics of the study population.

	Total (n = 57,820)	Women (n = 27,608)	Men (n = 29,923)	p value
Age, mean (SD), 95% CI	50.5 (11.8), 50.4–50.6	51.0 (11.5), 50.9–51.2	50.1 (12.1), 50.0–50.2	<0.001a
Number of quit attempts, mean (SD), 95% CI	1.16 (0.46), 1.16–1.17	1.17 (0.47), 1.17–1.18	1.15 (0.44), 1.15–1.16	<0.001a

Area of residence (n = 57,291)	n (%)	n (%)	n (%)	p value
Urban	30,707 (53.6)	15,206 (55.3)	15,501 (52.1)
Semi-urban	6,319 (11.0)	3,068 (11.1)	3,251 (10.9)
Rural	20,265 (35.4)	9,248 (33.6)	11,017 (37.0)	<0.001b

Data on age and sex were available for 57,531 individuals; age and/or sex were missing for 289 individuals. Data on area of residence were missing for 529 individuals. Values are expressed as mean (standard deviation [SD]) with 95% confidence interval (CI), or as number (percentage), as appropriate.

^aMann–Whitney U test.

^bFisher's exact test.

Women undertook a higher number of successive quit attempts than men (p < 0.001; Table 2).

According to geographical distribution based on the health system classification¹⁴ pharmacotherapy use corresponded to 2.3% of the urban population, 2.8% of the semi-urban population and 2.6% of the rural population¹³ (Table 2). In rural areas, women undertook fewer quit attempts than men (p < 0.001). In addition, women treated with varenicline showed greater adherence (collection of three or more packs) than men (p < 0.001), whereas no significant sex-related differences were observed for bupropion (p = 0.094).

Regarding income level, 13.7% of dispensations corresponded to individuals classified as TSI-1 and 18.0% to TSI-2 (both predominantly retired populations). A total of 43.6% corresponded to TSI-3 (annual income below €18,000), 24.2% to TSI-4 (income between €18,000 and €100,000), and 0.4% to TSI-5 (income above €100,000). A higher proportion of women was observed in the TSI-3 group (47.1% vs. 40.1% in men), while a lower proportion was found in the TSI-4 group (19.1% vs. 29.2%), with statistically significant differences (p < 0.001).

As shown in Fig. 2, medication use was lower during the summer months and in December compared with the rest of the year. However, these differences did not reach statistical significance (summer: p = 0.579; December: p = 0.282).

Figure 2.

Distribution of the number of smoking cessation attempts (n = 67,255) by year and month. March 2020 marks the onset of the COVID-19 pandemic. Bupropion availability was limited between 2022 and the second half of 2024. Funding of cytisinicline began in February 2023. Varenicline was withdrawn from the market in mid-2021 and reintroduced in the second half of 2024.

Discussion

This study highlights the low uptake of pharmacotherapy for nicotine dependence in real-world clinical practice and provides insight into patterns of use and treatment completion within a publicly funded healthcare system. The cumulative incidence of pharmacotherapy-assisted smoking cessation attempts was 14.5%, which was higher than the 11.9% reported in a previous study conducted in a single health area.¹¹ Nevertheless, completion of longer-duration treatments was low.

These findings are consistent with previous results, showing that 77% of patients treated with varenicline and 54% of those treated with bupropion collected fewer packs than those recommended and covered by public funding.¹¹ Similarly, a study conducted in the Netherlands reported low adherence rates for these treatments (14% for varenicline and 33% for bupropion), although adherence improved when pharmacotherapy was publicly funded.¹⁸

In addition, 13% of individuals in the analysed cohort made more than one quit attempt. Evidence on whether switching pharmacotherapy in subsequent attempts increases the likelihood of success remains limited.19, 20

However, current recommendations support offering the most effective therapeutic option at each attempt,²⁰ prioritising those that are available and likely to ensure better adherence.²¹

As previously emphasised, pharmacological treatment should be co-administered with sustained cognitive–behavioural interventions, scheduled follow-up visits, and monitoring of both adherence and adverse effects. This comprehensive approach not only improves adherence to pharmacotherapy²² but also contributes to higher smoking abstinence rates.

In the context of the withdrawal and supply shortages of varenicline and bupropion between mid-2021 and 2024, cytisinicline became the most widely used pharmacological option following its public funding from 2023 onwards. Recent clinical trials have consolidated evidence regarding its efficacy and safety,²³ findings that have also been confirmed in real-world clinical practice studies.²⁴ Moreover, this therapeutic option has gained increasing relevance due to the lower frequency and severity of adverse effects.²⁵

Although varenicline has demonstrated higher comparative abstinence rates than the 25-day cytisinicline regimen (13.3% vs. 11.7% with biochemical verification²⁶; 32.5% vs. 23.1% based on self-reported abstinence²⁷), it is also associated with a higher burden of adverse effects and lower adherence rates.²⁵

This issue is particularly relevant given that, in routine clinical practice, individuals using pharmacological treatments for tobacco dependence often take them for shorter durations and at lower doses than recommended, thereby compromising their effectiveness.²⁸

In this context, public funding not only increases the number of quit attempts,18, 29 but is also associated with a higher proportion of individuals who successfully quit smoking.9, 21

Available evidence suggests that between 18 and 25 out of every 100 individuals treated with smoking cessation pharmacotherapy achieve abstinence.²⁵ Based on these estimates, and from a public health perspective, the results of the present study suggest that between 2315 and 3215 individuals may have successfully quit smoking. However, this approximation should be interpreted with caution, as individual-level clinical data on abstinence were not available, and its purpose is solely to provide a contextual estimate of the potential population-level impact of the intervention.

Analysis by area of residence showed a higher proportion of pharmacotherapy use in rural settings, although utilisation among women was lower in these areas. These findings may be related to the geographical characteristics and population dispersion of the healthcare system analysed.

Notably, women made a higher number of quit attempts and showed slightly better treatment completion, with statistically significant differences. In addition, analysis according to co-payment status suggests a lower socioeconomic level among women who smoke. These results reinforce the need to adapt smoking cessation interventions using a gender-sensitive approach,¹⁰ and to remove financial barriers to access to these treatments.17, 30 Improving the effectiveness and equity of this public health measure,⁷ supports consideration of full public coverage.⁶ In this regard, our findings contribute to understanding the reach and utilisation of publicly funded smoking cessation pharmacotherapy programmes, rather than their clinical effectiveness.

Limitations

This study has several limitations. The use of pharmacy dispensing data allows adherence to be estimated indirectly based on the number of medication packs dispensed, but does not confirm actual medication intake. Furthermore, although the methodology allows the assumption that patients met the clinical eligibility criteria required for reimbursement – given the clinical modelling necessary for electronic prescribing – it does not allow the evaluation of outcomes in terms of smoking abstinence. Finally, temporal heterogeneity in the availability of different pharmacological agents limited the feasibility of multivariable analyses.

Conclusions

The analysis of pharmacy dispensing data provides a feasible approach to evaluating the implementation of publicly funded smoking cessation programmes in real-world clinical practice. The findings show limited utilisation and low completion of recommended treatment regimens, as well as differences according to sex and socioeconomic status. However, these findings should be interpreted with caution, as clinical outcomes in terms of smoking abstinence could not be assessed. In this context, the results primarily reflect patterns of use and implementation rather than the effectiveness of the interventions. Overall, the results highlight the need to optimise support strategies for individuals motivated to quit smoking, including adequate follow-up and removal of barriers to access, while also improving accessibility and equity in publicly funded smoking cessation programmes and advancing the integration of information systems to enable evaluation of their clinical impact.

What is known about the topic

• •Pharmacological treatment combined with behavioural support improves smoking cessation outcomes.
• •Adherence to smoking cessation pharmacotherapy is frequently suboptimal in routine clinical practice.
• •Public funding increases access to smoking cessation treatments.

What does this study adds

• •Use of funded smoking cessation pharmacotherapy was low in a public health system.
• •Adherence to recommended treatment regimens was limited, particularly for longer-duration therapies.
• •Differences in utilisation and adherence according to sex and socioeconomic status highlight persistent barriers in real-world access to funded smoking cessation treatments.

Authorship

RMG: Conceptualization, Methodology, Investigation, Supervision, Formal analysis, Writing – original draft, Writing – review & editing, Funding acquisition. ADM: Investigation, Supervision, Writing – review & editing, Funding acquisition. CDF: Investigation, Data curation, Writing – review & editing. SMMT: Investigation, Data curation, Writing – review & editing. MNF-M: Methodology, Formal analysis, Validation, Writing – original draft, Writing – review & editing. DF-G: Methodology, Supervision, Data curation, Formal analysis, Validation, Writing – original draft, Writing – review & editing.

Ethical considerations

The study was approved by the Research Ethics Committee with Medicines of the Zamora Healthcare Area in 2024 (registration number 657). Given the retrospective design and the use of previously anonymised data, the requirement for informed consent was waived.

Declaration of generative AI and AI-assisted technologies in the writing process

None of the materials has been produced partially or totally with the help of any artificial intelligence software or tool.

Funding

This work has been selected by the Gerencia Regional de Salud (SACYL), Consejería de Sanidad de la Junta de Castilla y León, with funding granted as part of a bio sanitary research, health management and social and health care project to be developed in 2025. GRS 2989/C/2024. Principal Investigator: Mr. Alfonso Díaz Madero.

Conflicts of interest

Raúl Majo García has received speaker fees from Chiesi and Aflofarm, paid expert testimony, participation in clinical studies and funding to attend conferences from Adamed. Sheila María Martínez Tahoces has received speaker fees from GlaxoSmithKline (GSK) for scientific and educational lectures. The other authors declare not to have any conflicts of interest that may be considered to influence directly or indirectly the content of the manuscript.