Light smokers are less likely to receive advice to quit from their GP than moderate-to-heavy smokers: A comparison of national survey data from the Netherlands and England

Abstract

Background: A substantial proportion of smokers nowadays smoke fewer than 10 cigarettes per day (cpd). These ‘light’ smokers are still exposed to significant health risks but may not receive as much attention from health professionals as heavier smokers. The Netherlands and England are two countries with very different smoking cessation treatment climates and may vary in the extent to which GPs advise on smoking cessation.

Objectives: To assess whether Dutch and English light smokers (< 10 cpd) are less likely to receive advice to stop smoking during a consultation with their GP than moderate-to-heavy smokers (≥ 10 cpd).

Methods: We compared data from two series of national surveys: the Dutch ‘Continuous Survey of Smoking Habits’ and the English ‘Smoking Toolkit Study’. We included respondents to both surveys from February 2010 to December 2011, aged 16+ years, who consulted their GP in the previous 12 months.

Results: A total of 7734 smokers responded to the surveys in the Netherlands and 10 383 in England. The percentage of Dutch smokers who received advice to quit from their GP was 22.6% (95% CI = 21.5–23.7) compared with 58.9% (95% CI = 57.6–60.2) of English smokers. Light smokers were less likely to receive advice to quit than moderate-to-heavy smokers, both in the Netherlands (OR = 0.57, 95% CI = 0.50–0.65) and in England (OR = 0.64, 95% CI = 0.57–0.72).

Conclusion: Smokers in the Netherlands are less than half as likely to receive advice to quit from their GP as smokers in England. In both countries, light smokers are less likely to receive advice to quit from their GP than moderate-to-heavy smokers.

Keywords::

• Smoking cessation
• general practitioners
• health surveys

KEY MESSAGE:

• Smokers in the Netherlands are less than half as likely to receive advice to quit from their GP as smokers in England.

• In the Netherlands and in England, light smokers are less likely to receive advice to quit from their GP than moderate- to-heavy smokers.

INTRODUCTION

The proportion of smokers who smoke less than an average of 10 cigarettes per day (cpd) has increased in Western countries during recent years (1,2). In the Netherlands and in England, these ‘light’ smokers now constitute approximately one third of current smokers (1, 3).

Light smoking poses a substantial health risk. Light and non-daily smoking carries nearly the same risk for cardiovascular disease as moderate-to-heavy smoking (4). The risk of lung and other cancers is lower than in daily smokers but substantially higher compared with non-smokers (4). Furthermore, light smokers are at increased risk of respiratory, reproductive, and other diseases, and face a higher all-cause mortality than non-smokers (4).

According to clinical guidelines, health care professionals should advise every smoking patient to stop smoking (5–7). Physician advice is generally effective in increasing attempts to quit and achieving long-term abstinence (8,9). However, little is known about the specific association between physician advice and level of smoking. One US population study from the year 2000 compared non-daily with daily smokers and found that non-daily smokers were 35% less likely to be asked by their physician about smoking and advised to quit (10).

The degree to which smokers receive advice and assistance to stop by their health care provider is likely to be country specific. For example, compared with Australia, Canada and the US, smokers in the UK visiting a physician or other health professional are much more likely to receive additional help or referral to a cessation service (11). A recent international study suggests that the rate at which smokers visit a doctor or other health professional is comparable, but among those who consulted a health professional, only 5% of smokers in the Netherlands were advised to quit compared with 28% in the UK (12).

England and the Netherlands have very different smoking cessation treatment climates. England has an infrastructure of smoking cessation clinics throughout the country where smokers are offered free behavioural support to stop smoking and free or extremely cheap medication (13). The availability of these clinics for referral of patients is likely to make it easier for GPs to address smoking. The Netherlands is characterized by a more decentralized free-market system. Private health insurers decide which treatment will be reimbursed and many parties offer smoking cessation programmes in an unregulated manner with regard to whether treatment is fully evidence-based. However, the government tries to improve the overall quality of services by advising health insurers to reimburse cessation services that adhere to specific quality criteria that are based on the national clinical guidelines for smoking cessation (14–16).

It is important to know whether light smokers are just as likely to receive advice to stop smoking from their GP as heavier smokers because the proportion of light smokers among current smokers is increasing and stopping smoking is still important for light smokers to reduce health risks. Furthermore, it would be interesting to know whether this likelihood differs between England and the Netherlands. We, therefore, conducted a study with national survey data from these two countries to assess whether light smokers (< 10 cpd) are less likely to receive advice to stop smoking during a consultation with their GP than moderate-to-heavy smokers (≥ 10 cpd), and whether the extent to which GPs offer advice differs between the countries. Furthermore, we assessed differences in GP’s prescription and recommendation of smoking cessation counselling and medication.

METHODS

We used data from two series of monthly national surveys from the Netherlands (the ‘Continuous Survey of Smoking Habits’, CSSH) (3) and England (the ‘Smoking Toolkit Study’, STS) (17). The CSSH is a weekly cross-sectional population survey that is used to monitor smoking habits of the Dutch population. It is conducted by TNS NIPO for the Dutch expert centre on tobacco control (STIVORO). Respondents for the CSSH were selected randomly from the TNS NIPObase, a database containing more than 140 000 potential respondents who participate in internet-based research on a regular basis. TNS NIPObase panel members are recruited actively by TNS NIPO using telephone and mail. The Central Committee on Research Involving Human Subjects in the Netherlands requires no ethical approval for non-medical survey research. The methods of the STS have been described in full elsewhere and have been shown to result in a sample that is nationally representative in its socio-demographic composition (17). In brief, each month a new random sample of approximately 1800 adults completes a face-to-face computer-assisted survey. The study was granted ethical approval by the University College London ethics committee.

For the present study, we used data from respondents to both surveys in the period from February 2010 through to December 2011 who were 16+ years of age, current smokers at the time of the survey, and provided data on the number of cigarettes per day and consultations with their GP in the 12 months prior to the survey.

Measurements

All respondents provided data on their age, sex, and cigarettes per day. We categorized respondents into light smokers (smoking on average < 10 cpd) and moderate-to-heavy smokers (≥ 10 cpd). Previous research showed that in England approximately three quarters of light smokers are daily smokers (1).

The level of tobacco dependence was measured with the variable ‘time to first cigarette’, which is a widely used and validated measure of dependence (18,19). Respondents were asked: ‘How soon after you wake up do you light up?’: within 5 min (= highest level of dependence), 6–30 min, 31–60 min, more than 60 min (= lowest level).

Smokers who consulted their GP in the past 12 months were asked whether smoking was discussed and whether they received advice to quit from their GP. Those who discussed smoking were asked whether their GP recommended or prescribed any evidence-based smoking cessation mediation or behavioural support. Details on variables measuring advice and support in each survey are provided in the Appendix.

Statistical analyses

Differences in baseline characteristics between light and moderate-to-heavy smokers were statistically tested with Chi-square tests for categorical variables and independent t-tests for continuous variables.

We used simple and adjusted logistic regression models with type of smoker as the independent variable (light vs. moderate-to-heavy). Separate models were used for the following dependent variables: consulted GP in the last year (yes vs. no); smoking discussed during consultation (yes vs. no); received advice to quit (yes vs. no); medication recommended/prescribed (yes vs. no); counselling recommended/prescribed (yes vs. no). Adjusted models included the variables age, sex, time to first cigarette, and month of the survey.

We used weighted data only for analyses in which population percentages were being estimated: the rate of light smokers and the rate of smokers who consulted their GP among all current smokers, as well as the rate of smokers who received advice to quit among those who consulted their GP. In the CSSH, data were weighted by region, urbanization, age, sex, household composition, and level of education. In the STS, data were weighted by age, sex, and socioeconomic group to match the 2001 Census.

We did not impute any missing data. Cases with missing data were excluded analysis-by-analysis (i.e. only in those analyses where they had missing data on a specific variable).

RESULTS

The number of current smokers who responded to the surveys was 7734 in the Netherlands and 10 383 in England. The weighted percentage of smokers who were classified as light smokers was lower in the Netherlands than in England: 29.3% (95% confidence interval (CI) = 28.3–30.3) compared with 33.0% (95% CI = 32.0–33.9).

Table 1 shows the differences in characteristics between light and moderate-to-heavy smokers in each country. Light smokers were more often female, younger, and reported lower levels of tobacco dependence. The age, sex and cpd of light and moderate-to-heavy smokers were similar between the two countries. Smokers in England, however, appeared to have higher levels of tobacco dependence.

Table 1. Differences in characteristics between light and moderate-to-heavy smokers, stratified by country.

	Netherlands			England
Characteristic	Light smokers % (n)	Moderate-to-heavy smokers % (n)	P	Light smokers % (n)	Moderate-to-heavy smokers % (n)	P
Male sex	47.3 (1051)	50.5 (2782)	0.010	46.9 (1540)	52.1 (3700)	< 0.001
Age, mean (SD)	41.3 (16.9)	45.1 (14.7)	< 0.001	40.1 (17.3)	43.8 (16.3)	< 0.001
Cigarettes per day, mean (SD)	4.2 (2.5)	18.4 (7.7)		4.5 (2.3)	16.8 (7.5)	< 0.001
Time to first cigarette
Within 5 min	1.5 (34)	20.0 (668)	< 0.001	7.1 (231)	27.3 (1922)	< 0.001
6–30 min	12.1 (268)	52.6 (1802)	19.0 (618)	40.3 (2851)
31–60 min	14.7 (328)	17.7 (599)	18.5 (509)	19.1 (1354)
More than 60 min	67.1 (1493)	9.3 (321)	55.4 (1,797)	13.3 (942)

Data are presented as column percentage (number), unless stated otherwise. Light smokers = smoking less than 10 cigarettes per day. Moderate-to-heavy smokers = smoking 10 or more cigarettes per day.

Consultation with GP

The number of current smokers who consulted their GP in the last year was 5488 in the Netherlands and 6155 in England. The weighted percentage of smokers consulting their GP was higher in the Netherlands than in England: 70.7% (95% CI = 69.7–71.7) compared with 57.6% (95% CI = 56.6–58.6).

Compared with moderate-to-heavy smokers, the odds of Dutch and English light smokers consulting their GP in the last year were 9% and 21% lower (as indicated by the unadjusted odds ratios presented in Table 2). These differences between levels of smoking dissipated largely when adjusted for age, sex, and tobacco dependence.

Table 2. Percentage light versus moderate-to-heavy smokers consulting their GP in the last year and receiving advice and support to quit, stratified by country.

	Netherlands				England
	Light smokers (%)	Moderate-to-heavy smokers (%)	OR (95% CI)	aOR (95% CI)	Light smokers (%)	Moderate-to-heavy smokers (%)	OR (95% CI)	aOR (95% CI)
In all smokers
Consulted their GP last year	69.9	72.0	0.91	1.12	55.4	61.1	0.79	0.91
			(0.81–1.01)	(0.98–1.29)			(0.73–0.86)	(0.83–0.99)
In smokers who consulted their GP in the last year
Smoking discussed during consultation	21.5	32.5	0.57	0.83	61.9	71.7	0.64	0.78
			(0.50–0.65)	(0.70–0.99)			(0.57–0.72)	(0.68–0.89)
Received advice to quit smoking	15.2	26.2	0.50	0.76	54.1	63.1	0.69	0.82
			(0.43–0.59)	(0.63–0.92)			(0.62–0.77)	(0.72–0.92)
In smokers who consulted their GP in the last year and discussed smoking
Recommended/prescribed medication	21.1	27.5	0.71	0.87	21.9	24.8	0.85	0.96
			(0.53–0.95)	(0.61–1.23)			(0.73–1.01)	(0.80–1.15)
Recommended/prescribed counselling	3.6	2.5	1.47	2.34	39.3	44.1	0.82	0.85
			(0.75–2.88)	(0.99–5.49)			(0.72–0.95)	(0.73–0.99)

Data are presented as column percentages. Light smokers = smoking less than 10 cigarettes per day. Moderate-to-heavy smokers = smoking 10 or more cigarettes per day. OR = odds ratio for light smokers compared with moderate-to-heavy smokers within each country stratum (respondents with missing data excluded analysis-by-analysis). aOR = odds ratio adjusted for sex, age, time to first cigarette and month of the survey.

Discussion of smoking and advice to quit

The number of current smokers who consulted their GP in the last year and discussed smoking was 1611 in the Netherlands and 4235 in England. The weighted percentage of Dutch smokers who received advice to quit from their GP during a consultation was about half of that of English smokers: 22.6% (95% CI = 21.5–23.7) compared with 58.9% (95% CI = 57.6–60.2).

Light smokers were less likely to discuss smoking during a consultation with their GP than moderate-to-heavy smokers: the odds were 43% lower in Dutch light smokers and 36% lower in English light smokers (Table 2). Light smokers were also less likely to receive advice to quit from their GP than moderate-to-heavy smokers: the odds were 50% lower in Dutch light smokers and 31% lower in English light smokers (Table 2). The lower odds in light smokers to discuss smoking and receive advice remained after adjustment.

Prescription/recommendation of evidence-based smoking cessation aids

The weighted percentage of light smokers who received a recommendation or prescription for smoking cessation medication from their GP was similar in the Netherlands and in England: 20.9% (95% CI = 16.4–25.3) and 23.6% (95% CI = 22.2–24.9), respectively. The weighted percentage of light smokers who received a recommendation or prescription for smoking cessation counselling from their GP was about one tenth in the Netherlands compared with England: 3.7% (95% CI = 1.7–5.8) and 42.7% (95% CI = 41.1–44.3), respectively.

The odds of light smokers in the Netherlands receiving a recommendation or prescription for smoking cessation medication from their GP were 29% lower than moderate-to-heavy smokers (Table 2). In England, light smokers were less likely to receive medication, but this was not statistically significant. In the Netherlands, there was no statistically significant difference between light and moderate-to-heavy smokers in the odds of receiving a recommendation or prescription for smoking cessation counselling, whereas the odds were 18% lower in English light smokers (Table 2).

DISCUSSION

Our comparison of population data suggests that Dutch smokers are less than half as likely as English smokers to receive advice to quit from their GP during a consultation. In both countries, light smokers are less likely to receive advice to quit than moderate-to-heavy smokers.

Strengths and limitations of the present study

The comparison of data from the Netherlands and England in our study is limited by the fact that we used data from two different surveys with different methodologies. For example, the CSSH is an internet-based study, whereas the STS is a household survey with face-to-face interviews. Furthermore, the wording of some of the items is slightly different (see Appendix). Therefore, we cannot rule out that part of our observed differences between countries (e.g. regarding the association between smoking and advice) may be due to differences in methodology and wording of items. Future research should cross-validate the measures and methodologies in the different populations, and without this validation direct comparison of data from the two surveys may be biased.

The results in this study are based on self-reported data collected from smokers. In an ideal situation, GP behaviour would be assessed with objective measures such as direct observations, but it is not feasible to conduct such a study on a national scale.

Another potential limitation of our data is recall bias. Respondents to the surveys were asked about a consultation with their GP during the last 12 months. It may be that respondents do not always recall exactly what was discussed during that consultation, especially when it was long ago, which would have led to under-reporting of GP advice and support in our study. Furthermore, a previous study has shown that patients’ recall of stop-smoking advice is related to having a smoking- related disease (20), which would have led to an overestimation of the relative difference in advice between light and moderate-to-heavy smokers. We have no reason to assume, however, that such bias was different in the two countries.

Our study aimed at assessing differences in prevalence of GP advice in light (< 10 cpd) versus moderate-to-heavy smokers (≥ 10 cpd). The underlying rationale was that light smoking poses a substantial health risk, and therefore, light smokers should be just as likely to receive advice to stop smoking from their GP as moderate-to-heavy smokers. We do not know, however, whether GP advice to stop is effective in light smokers. As far as we know, there have been no studies specifically looking at the efficacy of physician advice in light smokers. The Cochrane review on physician advice for smoking cessation includes 41 trials (8). The majority of these did not have an inclusion criterion beyond ‘smoker’, and given the prevalence of light smoking, their analyses will have included many light smokers. However, it appears that the authors of the original studies did not report the results of their trial for the different subgroups of smokers.

A further point of discussion is the stability of light smoking. Our classification of light smoking was only a snapshot and does not necessarily represent a stable behaviour pattern. It has been shown that a substantial minority of smokers are in transition between lower and higher levels of smoking (1). Such smokers may include, for example, young people who begin to smoke or older adults who try to quit (21). We used the cut-off point of 10 cpd because this cut-off point has been used frequently in previous research to distinguish lower-from higher-level smokers and in randomized controlled trials as a criterion of in-/exclusion.

The major advantage of our study is that we were able to compare data over a period of almost two years from two surveys using representative samples of the Dutch and English population. These data include detailed information about national smoking patterns and cessation-related behaviour as the surveys were designed for an up-to-date evaluation of tobacco control strategies.

Interpretation of the results in relation to existing literature

This is the first study to assess and compare GP advice and support for smoking cessation between subgroups of smokers (light versus moderate-to-heavy) and between countries (Netherlands versus England). The first important finding is that only 23% of smokers in the Netherlands receive advice to quit from their GP during a consultation compared with 59% of smokers in England. Our finding supports a recent international study showing that smokers in the Netherlands are much less likely to receive advice to quit than smokers in the UK (12). There may be several explanations for this observation. First, GPs in England are incentivized for registering their patients’ smoking status and for offering advice to quit smoking within the Quality and Outcomes Framework, which is part of the National Health Service’s (NHS) General Medical Services Contract. Second, there is a very good infrastructure available in England that enables GPs to refer their smoking patients to expert help for smoking cessation (the NHS Stop Smoking Services). The Netherland has a system dominated by free-market and a strong reliance on health insurance companies to decide on the type of treatment. However, more research is needed to link such infrastructural differences to actual GP advice and prescribing behaviour. Third, smoking cessation treatments in England are reimbursed. In the Netherlands, pharmacotherapy has not been reimbursed, except for the year 2011 in which the Dutch Government decided to make it part of the basic health insurance coverage, but this only lasted for 12 months. Furthermore, the level at which behavioural support is reimbursed in the Netherlands varies across health insurers and is not very transparent to patients, nor to physicians. Finally, cultural differences may play a role as well. The Dutch nation has a particular strong sense for personal liberty which is reflected in a national survey among a representative sample of Dutch GPs of which 93% agreed that it is the patient’s responsibility to decide if he/she stops smoking (22).

A second finding is that light smokers in England and the Netherlands are much less likely to receive advice to quit from their GP than moderate-to-heavy smokers. We found only one other study about this topic in the literature, conducted in the US, which reported a similar result: the odds of non-daily smokers receiving advice to quit from their physician were 35% lower than daily smokers (10). Possible explanations for this observation are that either light smokers present themselves to their GP less frequently with smoking-related complaints and diseases which trigger the GP to ask and advice about smoking (23), or that GPs pay less attention to the long-term health risks of low-level smoking.

Third, light smokers in England and the Netherlands are somewhat less likely to receive a recommendation or prescription for smoking cessation medication. This is not surprising given the fact that clinical guidelines recommend physicians to prescribe such medication only to smokers who are smoking at least 10 cpd because the relevant evidence is based on trials that usually exclude light smokers. Nevertheless, about one in five light smokers who discussed smoking with their GP received a recommendation or prescription for smoking cessation medication. This may seem low, but not every smoker who discusses smoking with his/her GP has the desire and intention to make an attempt to quit smoking following that discussion.

A fourth and striking finding is that only a very small fraction of smokers in the Netherlands receive a recommendation or prescription for smoking cessation counselling. Counselling is an essential component of effective smoking cessation treatment and recommended in clinical guidelines. Our finding that more than 95% of Dutch smokers are not referred to counselling leads to the assumption that the vast majority of smokers in the Netherlands do not receive the best care for the treatment of their addiction.

Implications for clinical practice

GPs should be made aware that light smokers nowadays form one third of the current smoking population. These smokers ought to receive the same level of advice and evidence-based assistance to stop smoking as moderate-to-heavy smokers, given the substantial health risks that are associated with light smoking. The low rate at which smokers are routinely advised to quit smoking and the low level of behavioural support they are being offered in Dutch general practice are alarming. There is an urgent need to promote tobacco control in Dutch general practice which may be achieved by a well implemented and transparent reimbursement scheme for pharmacological and behavioural treatments for smoking cessation and by introducing a nationwide smoking cessation infrastructure and referral system.

Conclusion

Smokers in the Netherlands are less than half as likely to receive advice to quit from their GP as smokers in England. In both countries, light smokers are less likely to receive advice to quit from their GP than moderate-to-heavy smokers.

ACKNOWLEDGEMENT

The authors thank Gera Nagelhout for compiling the CSSH dataset for this study.

FUNDING

The Dutch Continuous Survey of Smoking Habits is supported by grants from the Dutch Ministry of Health, Welfare and Sport. The Smoking Toolkit Study is funded by the English Department of Health, Cancer Research UK, Pfizer, GlaxoSmithKline, and Johnson & Johnson. Pfizer, Johnson & Johnson, and GlaxoSmithKline are manufacturers of smoking cessation products who had no involvement in the design of the study, collection, analysis or interpretation of the data, the writing of the report, or the decision to submit the paper for publication.

Declaration of interest: Robert West undertakes research and consultancy for, and has received travel expenses and hospitality from, companies that develop and market smoking cessation medications. He has a share in a patent for a novel nicotine delivery device. Daniel Kotz, Marc Willemsen, and Jamie Brown do not have a conflict of interest.

Appendix 1. Details about measurement of variables in the surveys.

Table

	Continuous Survey of Smoking Habits (CSSH)	Smoking Toolkit Study (STS)
Smoker	Respondents were asked: ‘Do you smoke now and then or do you smoke not at all?’:	Respondents were asked: ‘Which of the following best applies to you?’:
	(1) smokes (now and then);	(1) I smoke cigarettes (including hand-rolled) every day;
	(2) smokes not at all;	(2) I smoke cigarettes (including hand-rolled), but not every day;
	(9) don’t know. Respondents choosing response category 1 were defined as smokers.	(3) I do not smoke cigarettes at all, but I do smoke tobacco of some kind (e.g. pipe or cigar);
		(4) I have stopped smoking completely in the last year;
		(5) I stopped smoking completely more than a year ago;
		(6) I have never been a smoker (i.e., smoked for a year or more);
		(7) don’t know;
		(8) not stated.
		Respondents choosing any response category from 1 through to 3 were defined as smokers.
Advice and support to quit	Respondents were asked: ‘Have you consulted your GP in the past 12 months? (variable ‘consulted GP in the last year’: ‘yes’ versus ‘no’). Respondents who consulted their GP were asked: ‘Was the topic of smoking brought up during the consultation with the GP?’ (variable ‘smoking discussed during consultation’: ‘yes’ versus ‘no’) and ‘Did the GP say that you should stop smoking?’ (variable ‘received advice to quit’: ‘yes’ versus ‘no’). The following multiple choice question was asked about smoking cessation aids: ‘During the discussion about smoking (or smoking cessation), did your GP refer to the following aids?’: (1) nicotine microtab; (2) nicotine gum; (3) nicotine patch; (4) nicotine tablet; (5) Zyban/bupropion; (6) Notrilen/nortriptyline; (7) Champix/varenicline; (8) group behavioural support; (9) telephone quit line; (10) individual support by a nurse, behavioural therapist, or psychologist. In respondents choosing any response category from 1 through to and including 7, the variable ‘medication recommended/prescribed’ was coded ‘yes’ (otherwise ‘no’). In respondents choosing any response category from 8 through to and including 10, the variable ‘counselling recommended/prescribed’ was coded ‘yes’(otherwise ‘no’). 1 or 2, the variable ‘counselling recommended/prescribed’ was coded ‘yes’ (otherwise ‘no’).	Respondents were asked the following multiple choice question: ‘Has your GP spoken to you about smoking in the past year (i.e. last 12 months)?’: (1) yes, he/she suggested that I go to a specialist stop smoking advisor or group; (2) yes, he/she suggested that I see a nurse in the practice; (3) yes, he/she offered me a prescription for Champix, Zyban, a nicotine patch, nicotine gum or another nicotine product; (4) yes, he/she advised me to stop but did not offer anything; (5) yes, he/she asked me about my smoking but did not advise me to stop smoking; (6) no, I have seen my GP in the last year but he/she has not spoken to me about smoking; (7) no, I have not seen my GP in the last year. In respondents choosing any response category from 1 through to and including 6, the variable ‘consulted GP in the last year’ was coded ‘yes’ (otherwise ‘no’). In respondents choosing any response category from 1 through to and including 5, the variable ‘smoking discussed during consultation’ was coded ‘yes’ (otherwise ‘no’). In respondents choosing any response category from 1 through to and including 4, the variable ‘received advice to quit’ was coded ‘yes’ (otherwise ‘no’). In respondents choosing response category 3, the variable ‘medication recommended/prescribed’ was coded ‘yes’ (otherwise ‘no’).