Effectiveness of a stepped primary care smoking cessation intervention (ISTAPS study): design of a cluster randomised trial

The data from the 50-year follow-up study involving British male doctors showed that one in two regular smokers of cigarettes die as a result of tobacco-related disease. The latest estimations showed that in Spain in 2001 there were 54,233 deaths attributable to tobacco consumption [1]. The data of the National Health Survey [Encuesta Nacional de Salud] of 2006 showed that 29,9% of the adult Spanish population smoked [2], with a clear tendency towards a decrease in males and an increase in females.

Primary care can play a key role in the control of tobacco consumption. The data from our country show that in many zones more than 75% of the population attend an outpatient consultation in the local Primary-Care Centre (PCC) at least once a year, and the average number of visits varies between 5 and 6 (data from the primary-care database of 2001; Sistema de Información de Atención Primaria, 2001). Spain has an almost universal coverage of the population by a public health system in which primary care is the first point of contact. This provides a unique opportunity for intervention, and therefore prevention, in a large number of subjects on many occasions even before chronic or severe diseases have begun. Furthermore, surveys as those cited above showed that more than 60% of smokers had a desire to quit smoking, and the majority of these individuals have at some time or other made an attempt to achieve this.

In the past few years, several national and international clinical practical guidelines have been developed with recommendations to intervene on the consumption of tobacco, based on help provided at health centers, both at primary and specialist care levels.

Since 1996, various clinical guidelines integrating all the evidence available from existing clinical research have been published in different countries. Most of them were from anglosaxon countries (mainly the USA, UK and Australia). For example, the guidelines from the US Public Health services published in 2000 (recently updated)[3] were based on approximately 6000 original articles and abstracts of congresses written in English and published between 1977 and 1999. More than 50 meta-analysis were carried out and became the basis for the recommendations developed by a panel of 18 experts. Following this, there was an external revision process by another group of 70 experts [4].

The guidelines from the United Kingdom originated from the meta-analysis derived from the American guidelines and from the Cochrane international collaboration reviews [5]. Furthermore, in the year 2000, appeared the First European Directives on the Treatment of Tobacco Dependence within the European Cooperative Project of the World Health Organization.

Some of the recommendations of these guidelines have been based on a large number of clinical trials and studies of intervention. However, in other cases, there has been weak evidence and the recommendations have been based on the consensus opinions of experts.

Examples of intervention from the primary-care perspective for which there has been grade A evidence (based on many high quality clinical trials containing consistent results) are:

• Primary health care teams should ensure that their records concerning which of their patients smoke are kept up to date [A]

• GPs should advise current smokers to stop during routine consultations at least once a year, offer a prescription for Nicotine Replacement Therapy (NRT), record the response to that advice, and arrange follow up where appropriate. [A]

• At the beginning of the study, and even now, the level of evidence for some key recommendation were still C and mostly derived from published guidelines (recommendations based on consensus). These include:

▪ The role of motivational interviewing in accelerating changes in the process of smoking cessation.

▪ The role of the primary care professionals in performing more intensive interventions.

▪ The role of primary care nurses in motivating and helping patients to quit smoking.

▪ The use in primary care of non-nicotine based pharmaceutical products such as bupropion, an atypical antidepressant considered in the guidelines as a first-line drug. At the start of this study there was only limited experience of its use in primary care.

There is only partial evidence for several of these recommendations. For example, in the case of motivational interviewing [6, 7], the study by Butler et al showed outcomes less positive than those expected [8], and subsequent reviews showed some inconsistencies [9‐12]. Motivational interviewing is difficult to evaluate because of its intrinsic nature (variability and adaptation to characteristics both from patients and health professionals, and from the interaction between them in each clinical encounter) and also due to the difficulty of finding useful and appropiate measures of success. Another important aspect to consider is the fact that motivational interviewing requires a change in the way health professionals interact with patients.

The Cochrane review of smoking cessation interventions conducted by nurses showed that there can lead to potential benefits and that there is fair evidence of their effectiveness. Again there are few studies in primary care and some of these studies show that the interventions carried out during preventive check-ups are less effective [13].

The use of bupropion was not yet protocolised in primary care, although the Cochrane review does recommend it as a first-line drug. [14].

One of the conceptual models that has been most extensively used to explain the process of change in addictive behaviour is that developed by Prochaska and DiClemente in 1982. These researchers defined a series of successive stages each of which require different types of interventions [15, 16]. The use of this model is frequently linked to the motivational interviewing strategies described earlier. It is recommended that the interventions applied in primary care must be orientated towards achieving advances in this addiction process, and not only to giving up smoking. As such, stage change can be used as an intermediate measure of the effectiveness of the interventions. The majority of the clinical guidelines used in our Spanish environment include an assessment of the stage of smoking cessation in which the individual smoker is currently located [17].

Although the amount of research about smoking intervention is impressive, many of the studies have been conducted in highly selected populations of smokers and there have not been any studies that have evaluated all the range of interventions developed, applying them in a stepped manner to non-selected smokers who consult primary care centers for other reasons. Interventions include brief motivational interviews for smokers at the precontemplation – contemplation stage, brief intervention for smokers in preparation-action who do not want help, intensive intervention with pharmacotherapy for smokers in preparation-action who want help, and reinforcing intervention in the maintenance stage.

In Spain, the first studies date from the end of the 1980s when a group of primary care professionals in Barcelona (including two of the researchers of this study; Cabezas C and Martin C) replicated the effectiveness of the brief intervention study of Russell et al in the UK in 1979 [18, 19].

Since then, and extending their field of study to include the primary care environment, the same research team have evaluated the effectiveness of nicotine chewing gum and nurse-based advice [20]. Subsequent studies by Quilez, Salvador, Comas, Torrecilla, etc. [21‐24] have found results consistent with those of the international studies. Some of these studies had also been carried out in primary care centers, but by health-care professionals specially trained in this area.

Among the published studies there are at least two types of interventions which have been widely evaluated: the Doctors Helping Smokers (DHS) Program proposed by Solberg [25, 26], adapted to Spain by Solberg and Nebot and applied in Spain by Martin, Casas, Cordoba and Ballvé (members of the ISTAPS research group) and the Program to Quit Tobacco by Grandes and Cortada [27]. The DHS program is based on the proposals of the first guidelines of American National Cancer Institute at the end of the 1980s which are easily applied in primary care. The main limitations of the program arose from the very brief intervention proposed and from the lack of a formal evaluation of its effectiveness in Spain. [28].

The study by Grandes et al. is a quasi-experimental study which included a large number of smokers with an evidence-based strategy. However, it does not include current recommended strategies such as motivational interviewing, more intensive help (the authors evaluated a therapeutic plan with three visits and one telephone call), nor the use of other drugs (other substitutes for nicotine than patches and bupropion) [29].

Hence, when faced with reviewing the evidence and developing a clinical guideline for Spain, the authors of the present project found that there was no clear evidence either for some of the proposed recommendations, or for the overall strategy. This led to the current proposition for a study identifying this evidence. In order to put into practice the intervention proposed health professionals must be trained. Some parts of the intervention (those related with motivational interviewing) require not only the development of skills but also a change in the way health professionals interact with patients. For this reason, and in order to minimize contamination, the unit of randomization and analysis was the basic care unit consisting of a physician and nurse who take care of the same list of patients, both of whom received a 20-hours training course at the beginning of the study.

The opportunity to detect and follow-up a cohort of persons from before quitting smoking to more than 1 years post-quitting would facilitate the study of variation in the overall health-related quality of life produced as a result of the process of quitting smoking, and also provide a measure of the impact on health status. The studies published about this theme are comparisons of the perceptions of health-related quality of life in smokers and ex-smokers [30]. Serial data in clinically relevant moments of the smoking cessation process would highlight the times and the scale at which the most relevant changes in the health-related quality of life are produced.

The SF-36 is a multi-purpose, short-form health survey validated in Spain by Alonso et al [31]. It consists of 36 questions measured in an ordinal scale. It can be self-administered or used in personal or telephone interviews. This yields a profile of functional health and well-being score, as well as psychometrically-based physical and mental health summary measures The profile considers 8 scales: Physical Functioning, Role-Physical, Bodily Pain, General Health, Vitality, Social Functioning, Role-Emotional, Mental Health and two summary measures: Physical Component Summary (PCS) and Mental Component Summary (MCS).

Additionally, one of the items evaluates the change in health status over the previous year. The items detect not only positive but also negative health status. The SF-36 is sufficiently sensitive to changes when compared to other questionnaires such as the Nothingham or the Sickness Impact Profile that have been especially useful in clinical trials up until now. One version of the SF36 that measures the health perception over the precedent week, known as the acute version, enables the changes to be evaluated within the interval of time envisaged in the present study.

The present study evaluates, under near real conditions, the effectiveness in primary care of an overall strategy of smoking cessation, that includes specific interventions addressed to people in all the stages-of-change. It is one of the biggest studies ever undertaken, not only in primary care but at all levels of clinical services. Given the almost universal coverage of primary care in Spain, and considering that many people in the country belongs to social class III to V, it is expected that the majority of the participants of the study will belong to these social classes, generating evidence of effectiveness in this type of population.

Another strength of the study is the large number of health-care professionals and patients who are involved, reinforcing the external validity of the data.

The theoretical framework used is the stages-of-change model of Prochaska and Di Clemente. Despite the fact that some authors such as West [33] or Riensma [12] have questioned the utility of the model in the past years, it remains as one of the most used conceptual frameworks. Since one of the criticism made of the studies using the model is the use of only one of the constructs (the stages of change), in our study we make use of all of them: stages, processes, self-efficacy and decisional balance; we also validate our Spanish version in a complementary study. Health-care professionals who volunteered to participate in the present study could have been more motivated than the health-care professionals working in other PCCs. If this were to be the case, any resulting bias would be in the direction of more conservative estimates of the effects of the intervention. It has been decided to randomize basic care unit instead of smokers as it would be very difficult for a health-care professional (particularly one who is well trained in applying a specific anti-smoking intervention) to withhold this intervention in accordance with the randomization procedure for specific individuals; this would be especially true in the case of the patient having certain pathologies related with smoking or those who had tried to quit many times before, etc. Given the feasibility of recruiting a considerable number of basic care units in control and intervention groups, we believe that this type of cluster randomized assignment is necessary and we hope this will result in comparable groups of basic care unit (health professionals) and patients.

Although the unit of randomization is the basic care unit, the intervention is given at the individual level and the outcomes are also measured at this level. The estimation of the design effect will be taken into acount in all the main analysis.

The ISTAPS study is intended to be a pragmatic trial, trying to develop an intervention that will be applied easily in routine clinical practice. For this reason patients from 14 to 75 years coming to the practice for any reason were included as it is unusual for patients in Spain to consult specifically for smoking cessation advice; in many ocasions these patients presented with comorbidity.

One of the unique roles of primary care teams in smoking cessation and control is to increase the motivation of smokers to quit. In fact, in the many years which elapse between the initiation of smoking (in Spain from 13 to 16 years old) until the beginning of smoking related diseases or other causes of medical consultations, primary care are the only services used by healthy and young or adult population. The ISTAPS study will include more than 60% of their patients in the precontemplation and contemplation stages (approx 1800) and that will allow us to study extensively the effect of motivational interviewing in order to enable an advance in the stage-of-change.

Since the practical clinical guidelines used in the present study have not been yet disseminated, it will enable us to perform the evaluation of the effectiveness using an experimental study design. The guideline will not be available until the end of 2008 and, as such, the study results will not have been influenced by the widespread diffusion of the guidelines.

The availability of health-related quality-of-life data enables us to evaluate the impact of quitting smoking on general health, as well as on specific domains explored by the SF-36 questionnaire. That will be of use when scheduling visits of patients to primary care centres in the post-cessation period, so decreasing the possibility to relapse.

This study aims to demonstrate the effectiveness of a smoking cessation intervention at the two year post-intervention follow-up. The results will be useful to primary care health professionals in their daily practice.