Smoking cessation in patients with diabetes mellitus: Results from a controlled study of an intervention programme in primary healthcare in Sweden

Abstract

Objective. To evaluate an intervention programme on smoking cessation in patients with diabetes mellitus in primary healthcare. Design. Regional controlled intervention study. Setting. Seventeen primary healthcare centres in Sweden. Intervention. In the intervention centres, nurses with education in diabetes were given one half-day of training in motivational interviewing and smoking cessation. An invitation to participate in a smoking cessation group was mailed to patients from the intervention centres followed by a telephone call from the patient's diabetes nurse. The nurses who intervened were specially educated in smoking cessation. The control group received a letter containing advice to stop smoking and information about a one-year follow-up. Patients. Daily smokers with diabetes mellitus, 30–75 years of age. In the intervention centres 241 patients fulfilled the criteria and in the control centres 171 patients. Main outcome measures. Self-reported smoking habits after one year. Results. In total, 21% of the smokers accepted group treatment. After 12 months, 20% (42/211) in the intervention centres reported that they had stopped smoking and 7% (10/140) in the control centres; 40% (19/47) of the smokers who had participated in group treatment reported that they had stopped smoking. Conclusion. A computerized record system for patients in primary healthcare was used to identify diabetic patients who were smokers. The selected group was invited to a stop smoking programme. At a one-year follow-up significantly more patients in the intervention centres had stopped smoking compared with patients in the control centres.

• Diabetes mellitus
• general practice
• intervention
• smoking cessation

Smoking is a risk factor for type 2 diabetes [1], and in patients with diabetes smoking is associated with increased levels of HbA1c and microalbuminuria according to data from the Swedish National Diabetes Register [2]. The Nurses Health Study showed that women with diabetes who smoke had mortality three to four times higher than women with diabetes who had never smoked [3]. In a meta-analysis of men with and without diabetes, Yudkin found that intervention on smoking was the best way to prolong life in patients with diabetes [4].

There are few studies on smoking cessation in patients with diabetes, most of these with rather pessimistic results [5–7]. The American Diabetes Association recommends intensified smoking cessation efforts for diabetes patients [8]. It is important to consider smoking, other behavioural factors, and quality of life in the risk balance sheet for diabetes patients and not to concentrate totally on pharmacological treatment [9], [10].

Smoking is the most important risk factor for a patient with diabetes mellitus concerning both morbidity and mortality.

• A computerized medical record system is a useful tool for quality improvement, by making it easy to select patient groups for preventive measures.

• An invitation to a smoking cessation group mailed to smoking diabetes patients, followed by a telephone call from a diabetes nurse, resulted in 21% (50/241) of the patients accepting the offer. After one year, 40% of these (19/47) reported that they had stopped smoking.

Primary healthcare has the main responsibility for treatment of patients with type 2 diabetes in Sweden. Almost all primary healthcare centres have one or two nurses with special education and training in diabetes care. Many primary health centres also have a nurse with education and training in smoking cessation. A computerized medical record with registration of diagnoses and smoking habits makes it easy to select a special group. This method was successfully used in the primary healthcare centre of Habo for selecting all diabetes patients registered as smokers and to send them a personal invitation to join a smoking cessation group [11].

The aim of this study was to investigate an intervention programme, in a controlled multi-centre study, where all smoking patients with diabetes mellitus, identified by the computerized medical records, were sent an offer to participate in a smoking cessation group.

Material and methods

An invitation to participate in the study as either an intervention centre or a control centre was sent to the head and to the diabetes nurses of all 104 primary healthcare centres in the region, consisting of three county councils in the south-east of Sweden. Eleven centres agreed to be an intervention centre and six to be a control centre. All of these centres were located in the counties of Jönköping and Östergötland. Four centres in Vetlanda and two centres (Vaggeryd-Skillingaryd) share the same diabetes nurses and were therefore grouped together. In the first phase of the study, both the intervention and control centres identified patients, 30–75 years of age, treated at the primary healthcare centre for diabetes mellitus. The computerized medical records for these patients were checked for registration of smoking habits. If no such registration was present, the diabetes nurse contacted the patient for clarification. Patients were classified as smokers if they were daily smokers or had stopped smoking within the last month.

All smokers with diabetes from the intervention centres received an invitation letter to participate in a smoking cessation group. The letters were signed by their GP and diabetes nurse. The group sessions were free of charge. About one week later, the diabetes nurse called the patients to find out if they had accepted the offer to participate in the study, which consisted of one telephone interview and follow-up telephone calls after 6 and 12 months. The structured telephone interview consisted of questions about length and amount of their smoking and the Fagerstöm Test for Nicotine Dependence (six questions, 0 to 10 points) [12]. Other chronic diseases, e.g. cardiovascular disease, different kinds of treatment for diabetes, and duration of the diabetes disease, were also registered.

In the control centres, the patients with diabetes were sent a questionnaire with the same questions asked by the intervention centres. This letter also pointed out the risks of smoking and gave advice to stop smoking. The patients were also asked to participate in 12 months’ follow-up of their smoking habits. Only patients who agreed to participate in the study were contacted at the follow-up.

The diabetes nurses in the intervention centres were given one half-day of training in motivational interviewing and smoking cessation.

The intervention programme consisted of eight group sessions in a two-month period led by nurses with special education in smoking cessation. A group model for smoking cessation, described by Hjalmarson and recommended by the Swedish National Institute of Public Health, was used as a base for the smoking cessation groups [13]. Each group meeting lasted for 45–60 minutes. Issues discussed during the sessions were motivation to stop smoking, and advice on how to break the habit and how to prevent relapse. Pharmacological treatment, NRT (nicotine replacement therapy), or Bupropion was recommended. After the group treatment, the patients received individual support and follow-up by telephone calls 3, 6, and 12 months after the quit day.

Smokers in the intervention centres who did not want to participate in the group treatment received a telephone call after 6 and 12 months and were asked about their smoking habits.

After 12 months, the patients in the control centres were sent a questionnaire about their smoking habits. The patients who did not respond to the questionnaire were contacted by telephone.

The patient's self-declared amount of smoking was registered but no confirmation was made with biochemical methods. Seventeen patients (12 from the intervention centres and 5 from the control centres) died before the 12-month follow-up and were excluded from the analysis. Patients who did not respond to the 12-month follow-up were counted as smokers.

Ethics

The study was approved by the ethical committee of Linköping University.

Statistical methods

The proportion of participants in the different groups was compared with chi-square test using Yates correction or Fisher's exact test. A t-test was used to compare mean values.

Results

In the studied age groups, there were 1767 patients with diabetes mellitus in the intervention centres and 1124 in the control centres. The flow of the study is shown in Figure 1. In total, 241 (13.6%) of the diabetes patients were registered as smokers in the intervention centres and 171 (15.2%) in the control centres (Table I). The proportion of smokers varied between the different primary healthcare centres. In the intervention centres the range was 9.2% to 18.9%, and in the control centres 8.3% to 25.8%. Three centres, all of them intervention centres, had a complete registration of smoking habits in the medical records.

Figure 1.  Flow chart of the study.

Table I.  Total number of diabetes patients, number of diabetes patients for whom smoking habits were missing in the medical records, number of smokers, participants in the study, and percentage of smokers among the diabetes patients in the intervention and control centres respectively.

Centres	Number of diabetes patients	Smoking habits missing	Number of smokers	Participants in the study	Smokers (%)
Intervention centres
Ryd, Linköping	146	2	27	27	18.5
Åby	220	0	30	30	13.6
Valdemarsvik	239	0	22	24	9.2
Norrahammar, Jönköping	205	0	19	15	9.3
Vetlanda (4 centres)	607	132^1	83	82	13.7
Råslätt, Jönköping	227	13	43	36	18.9
Vaggeryd-Skillingaryd	123	4	17	9	13.8
Total	1767	151	241	223	13.6
Control centres
Väven, Norrköping	204	3	52	52	25.5
Kungsgatan, Linköping	132	30	34	32	25.8
Hälsan 2. Jönköping	129	16	17	16	13.2
Rosenlund, Jönköping	255	63	33	12	12.9
Bankeryd, Jönköping	253	16	21	21	8.3
Aneby	151	54^2	14	12	14.4
Total	1124	182	171	145	15.2

¹The smoking habits of most patients, but not all, were checked. It is not known how many were missed. ²Only patients known to be smokers were sent a letter. The smoking habits of the other patients were not checked.

Among those who fulfilled the criteria for inclusion in the study, 223 patients in the intervention centres and 145 patients in the control centres agreed to participate in the study. Baseline characteristics of the participants are given in Table II. There was no difference in baseline characteristics between participants in the intervention and control centres except for duration of smoking. Diabetes patients in the intervention centres had smoked for significantly more years than patients in the control centres.

Table II.  Baseline characteristics of the smoking diabetes patients in the intervention and control centres

	Intervention centres	Control centres
Number of women/men	98/125	60/85
Mean age (years)	59.4	60.6
Mean diabetes duration (years)	6.8	7.3
Treatment
Diet and exercise (%)	30.9	26.9
Tablets (%)	38.1	42.8
Tablets and insulin (%)	10.3	9.0
Insulin (%)	18.8	21.4
Cigarette smokers(%)^1	93.7	90.3
Mean duration of smoking (years)	39.2	36.4*
Mean value Fagerström test	3.9	3.9

¹The rest of the smokers were pipe smokers, or smoked cigars or cigar-cigarettes. *p < 0.05.

A total of 21% (50/241) of the smokers in the intervention centres accepted the offer of cessation treatment. In one of the intervention centres, none of the diabetes patients who were smokers accepted the offer of intervention.

After 12 months, 20% (42/211) in the intervention centres and 7% (10/140) in the control centres reported that they had stopped smoking. The difference was statistically significant (p < 0.01). Of those smokers who had participated in group treatment, 40% (19/47) reported that they had stopped smoking. Among those in the intervention group who did not take part in group treatment, 14% (23/164) reported that they had stopped smoking, a result significantly lower than for group treatment (p < 0.01).

Pharmacological treatment was used by 29 patients in the intervention group, 21 patients used NRT, and eight patients Bupropion. There was no significant difference in smoking cessation between the patients who had used pharmacological treatment and those who had not. Of those patients in the intervention centres who did not want to participate in group treatment, eight patients reported that they had used nicotine replacement therapy and one had used Bupropion.

Discussion

This study shows the opportunity in primary healthcare to use computerized medical records to select patients for preventive activities. Screening the records for smoking status also led to increased smoking cessation counselling for a selected group of smokers in another study [14]. In the intervention centres, 21% (50/241) of the smokers accepted the offer of cessation treatment; 40% (19/47) of these patients reported that they were non-smokers after one year. Even among those who did not accept group treatment, a significant number of diabetes patients reported that they had stopped smoking at the 12-month follow-up, compared with diabetes patients in the control centres. Before the study, all the intervention centres already had access to nurses specially trained in smoking cessation. In spite of this, the study shows that there is a potential for increasing smoking cessation among patients with diabetes mellitus. It seems that a letter with an offer to participate in a smoking cessation group free of charge, followed by a telephone call from the diabetes nurse, helped diabetes patients to stop smoking. It is also possible that the diabetes nurses taking part in a study, following smoking cessation education, could have had a more active attitude towards smokers than nurses in the control centres.

This study has some limitations. For practical reasons we had no biochemical verification of self-reported non-smoking. However, one conclusion from a Norwegian and Finnish study was that the deception rate in a similar group of patients was low [15], [16]. Other studies have shown some doubt about the value of biochemical verification. A Swedish study in primary healthcare showed a difference between plasma cotinine and nicotine levels [17]. Also, when using ordinary cut-off points for cotinine, a rather large number of never smokers were misclassified as smokers [18].

The centres in this study were not randomized. In an effort to give the centres an opportunity to calculate whether they had the available resources to participate as an intervention centre, the centres themselves chose to be either an intervention or a control centre. The percentage of smokers in the intervention and control conditions turned out to be about the same.

In the previously mentioned primary healthcare study in Habo, more than 60% in the intervention group reported that they were ex-smokers at 18-month follow-up [11]. However, this study was smaller, had no control group, and the patients had free access to NRT for three months.

The proportion of ex-smokers (40%) after 12 months among patients treated in groups is high in our study compared with other intervention studies concerning diabetes patients.

In a German study from a diabetes department clinic, 794 patients smoking more than five cigarettes per day were invited by mail to participate in a smoking cessation programme [5]. In total, 11% of the smokers agreed to participate and were randomized into intervention or control groups; 5% were non-smokers after six months in a structured behaviour therapy anti-smoking programme and 16% in the control group receiving anti-smoking advice given once by a physician. In a Spanish study, diabetes patients from both primary care and two hospitals were initially invited to participate in a general lifestyle study [19]. After a structured interview with a nurse, 280 patients were randomized to an experimental or control group. The intervention programme was based on individual contacts with a nurse including telephone contacts and letters. Transdermal NRT was recommended. After six months, 17% of the smoking diabetes patients in the intervention group had stopped smoking, compared with 2.3% in the usual care group. In both these studies biochemical verification was used.

In conclusion, our study has shown that an intervention programme in primary healthcare, by using the possibilities of computerized medical records, can help a significant number of patients with diabetes to stop smoking.

This study was supported by FORSS (Medical Research Council of Southeast Sweden), and the county councils of Jönköping and Östergötland. Professor Calle Bengtsson, Department of Primary Health Care Göteborg University, gave valuable comments on the manuscript. Special thanks are also offered to the nurses who participated in the study.