  |
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Journal of the New Zealand Medical Association,
02-May-2003, Vol 116 No 1173
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Improving the effectiveness of smoking cessation in primary
care: lessons learned
Dee Richards, Les Toop, Keith Brockway, Sue Graham, Bill
McSweeney, Donna MacLean, Margaret Sutherland and Alison Parsons
The serious health effects of smoking are indisputable.
Smoking is responsible for 20% of the deaths in most Western countries, shortens
the life expectancy of addicts by an average of eight years and adds a huge and
preventable burden of disease to over-stretched health
systems.1 Despite this, progress both in
primary prevention of addiction, using legislative and fiscal restrictions, and
secondary prevention, using smoking cessation programmes and population-based
interventions, has been slow.
Brief, opportunistic advice on stopping smoking and
non-tailored smoking cessation letters both increase cessation rates by
2–3%.2 There are a number of randomized
controlled trials supporting the effectiveness of nicotine replacement therapy
(NRT). A systematic review in 1994 found an overall, one-year quit rate of
15%.3
A meta-analysis in the same year found an overall quit rate of 22% at six
months. A recent Cochrane review of NRT
efficacy included studies with endpoints six months and beyond. An overall 14%
quit rate was calculated.4
A number of other interesting points were highlighted in
this review, including the fact that a key determinant of programme success is
the setting in which it is offered, with studies set in primary care showing
smaller effect than those in specialised clinics or studies using volunteers.
Suggested reasons for this were training differences, as well as the
often-encountered problems of translating research evidence into ‘real
world’ general practice – it was felt that differential rates
reflected the selection of motivated volunteers compared with the more
heterogeneous general practice population. This differential rate of success is
of some concern as the general practice sector would regard smoking cessation to
be one of its core functions, and delivery of smoking cessation programmes in
primary care has been shown to be cost
effective.5
There is a great deal of interest in addressing
cardiovascular risk factors in the primary care setting, and in the utility and
funding of smoking cessation and NRT in community settings. Despite evidence
that in New Zealand general practitioners (GPs) provide smoking cessation to
many patients, a recent study showed that New Zealand smokers are not well
informed about smoking cessation strategies and their
efficacy.6 A recent paper found differences
between actual and recommended practice in primary care in New Zealand and
identified a number of potential barriers to smoking cessation in primary care,
including time pressure and the fee-for-service
system.7 It was suggested that increased
practice nurse (PN) input could be useful.
It is important to assess the performance of a
well-supported, multifaceted, NRT programme when implemented in a general
primary care setting. Pegasus Health (PH) is an independent practitioner
association (IPA) situated in the Christchurch urban area that services a
primarily European population. It was formed in 1992, operates a collective
approach to GP budget holding and currently has over 225 GP and 240 associated
PN members. Pegasus Health has been running a smoking cessation programme for
some years (the PEGS programme: Preparation, Education, Giving up and Staying
smoke free). The programme was initially introduced as the ‘Smokescreen
Programme’ in both primary and secondary care. This was devised for use in
primary care in Australia8 and is based on the
‘readiness to change’ model9
(precontemplative/contemplative/ready), incorporating a nominated quit date and
NRT, and accompanied by supportive counselling and literature. An initial study
was performed after its introduction in 1995 using self-reported, six-month quit
rates and once again showed differential quit rates in primary and secondary
care, with rates of 17% in secondary care and 10% in primary
care,10 and a biochemically validated deception
rate using exhaled carbon monoxide of 14.4%. The primary care programme has been
substantially modified to make the programme more locally relevant and to try to
improve quit rates (the programme has been renamed the PEGS
programme).
The primary aim of this study was to estimate programme
uptake and six-month quit rates for patients enrolled in a New Zealand,
general-practice-based, smoking cessation programme. The programme combines NRT
with supportive counselling and has been implemented broadly in a primary care
setting in Christchurch, New Zealand.
The secondary aim was to use a wide range of patient,
practice and environmental variables to estimate any predictive effect on
outcome.
MethodsCurrent
intervention A half-time programme coordinator provided training,
materials and a readily accessible support service. Training and resources were
provided by the coordinator to practices on an individual basis that allowed the
programme to be tailored to each practice’s working style and patient
population. The training incorporated aspects of the model of behavioural
change, motivational interviewing, quitting strategies, NRT use, and relapse
prevention. The coordinator also provided ongoing updates, support and advice.
The IPA encouraged frequent (weekly to two-weekly) practice contact and PN
involvement was encouraged after the first consultation. One of the key
philosophies underpinning the implementation was to allow flexibility in
delivery between practices. Practices used the programme in a variety of ways,
from nurse-run clinics/groups to a GP/PN team approach. Each practice chose how
they wished to run the programme, usually based on the skill mix and interests
of GPs and PNs within the team.
While the criteria for patient enrolment remained the same as the original model,9 material for GPs and patients was rewritten and draws on the US Preventive Services Task Force’s Guide to clinical preventive services.11 Booklets were made shorter with less text, simple language, a ‘positive benefits of quitting’ perspective, and less emphasis on the health risks of smoking. There is less emphasis on the smoking cessation ‘battle’ and more on encouragement and support using a matter-of-fact tone. NRT is emphasised as central to the programme rather than just one option, with subsidisation of the NRT by Pegasus Health underpinning this. The ‘Quitters’ booklet is very patient interactive and used as a brief ‘workbook’ with emphasis on individualising the book to each smoker (eg, supports, reasons for and benefits from quitting, identification of likely relapse moods and events, nomination of rewards for quitting). Consultations covered, in no particular order, assessment of motivation to quit, nominated quit date, discussion of usage of NRT, use of motivational techniques, and discussion of behaviour changes. Patients contributed NZ$15 a week towards the cost of NRT as well as the initial consultation cost. Pegasus Health met the remaining costs of NRT supply. PN involvement was encouraged with flexible roles – either the GP or the PN may implement the programme depending on the availability and skill mix of the individual practice team. This is consistent with effective teamwork principles and was thought to offer added benefits for patient access by reducing costs, as the majority of practices in which PNs implement the programme can reduce or waive charges to the patient for follow-up visits. Data collection All patients enrolling in the programme provide information, which is recorded on an enrolment form, about their smoking history, previous quit attempts, and smoking-related diseases. A cohort of 516 patients enrolling in the programme over a two-month period (March and April 2000) were selected for study. The cohort were contacted by mail six months after their nominated quit date and asked to fill in a simple questionnaire. To maximise the response rate, those who did not respond were sent repeat questionnaires at least twice and then telephoned at least three times before being classified as non-contactable. Attempts were made to determine the new addresses of those who had moved since participating in the programme. The questionnaire gathered information about age, sex and ethnicity (using the New Zealand Census format). Address at enrolment was geo-coded and a New Zealand Index of Deprivation (NZDep96) score assigned as an indicator of socioeconomic status.12 At six months, patients were asked about their motivation for enrolling in the programme, where they had heard about the programme, whether they were currently smoking and, if they were, why they felt they had continued to smoke. Patients were also asked how long they had used NRT, whether they lived with other smokers, and who they had seen at the practice (primarily PN, GP or a combination). Responses to the questions about demographics, smoking and disease history were all categorical and pre-coded prior to study commencement. Responses to the questions about motivation for quitting, reasons for restarting, and suggestions for improvements were also categorical, with the categories developed from free-text answers during the first interviews, which were conducted by telephone to allow this. A free-text ‘other’ category allowed extra categories to be built in if necessary. Analysis Simple descriptive statistics were calculated in Excel. SAS (Version 8.02) and Confidence Interval Analysis (CIA)13 were used for the more detailed statistical analyses. ResultsA total of 3670 patients enrolled in
the programme in 2000 in 94 member practices (227 GPs). Of the 516 participants
enrolled in a consecutive two-month period, 334 (65%) were contacted by mail or
telephone. Over one third of participants had changed address in the six months
since participating in the programme so, as previously discussed, considerable
effort was required to achieve this response rate. The majority of
non-contactable participants had moved and no forwarding addresses were
available.
Table 1. Demographic characteristics of participants
(age, sex, ethnicity, NZDep96 score)
*there was a statistically significant difference in
response rates across different age bands: 51.89 degrees of freedom 6, p
<0.00001; †equivalent percentage of
Christchurch population
The majority of enrolled patients (57%) and respondents
(63%) were female. Almost all (92%) enrolled and respondent patients had smoked
for greater than five years and around two thirds had smoked for 15 years or
more. One quarter had a documented, smoking-related disease, and enrolled
patients and respondents had made on average two previous quit attempts. Eighty
eight per cent of respondents were European, 4% Maori, 4% other European, and 4%
‘other’. These small numbers did not allow any sub-analysis within
specific ethnic groups. (Demographic and socioeconomic data are shown in Tables
1 and 2).
Table 2. Socioeconomic distribution of programme
participants
*1 = least deprived, 10 = most deprived
Primary delivery role was split between PN (54%), GP (25%),
and combined delivery (16%). Patients heard about the programme mostly from
their GP (49%) or by word of mouth (34%). One fifth of respondents had used NRT
in previous quit attempts. Reasons for wishing to quit were varied – 38%
said they just did not want to be smokers any more, 19% said they enrolled in
the programme because of recent health deterioration, and 10% enrolled under
pressure from family and friends.
The overall six-month quit rate was 36% (95% CI
31–41). Assuming all non-contactable respondents are still smoking, a
‘worst-case-scenario’, six-month quit rate was calculated at 23%
(95% CI 20–27). This is likely to underestimate the true quit rate, as
most missing data were from untraceable patients who had moved rather than
non-responders. Using the deception rate of 14% established in the initial study
using exhaled carbon monoxide measurement10
this figure remains above 20%.
Univariate analysis initially showed duration of NRT (p =
0.03) and duration of NRT therapy (p = 0.03) were significant predictors of
quitting. Five to six weeks of NRT therapy was significantly better than one to
two weeks (OR 0.38, 95% CI 0.19–0.95), three to four weeks (OR 0.25, 95%
CI 0.06–0.98) or 11 to 12 weeks (OR 0.35, 95% CI 0.16–0.78). Age
group 45–55 years had a significantly better quit rate (50%) than
26–35 years, 36–45 years and 66–75 years. Conversely, living
with a smoker (p = 0.02), having made no previous quit attempts (p = 0.02) and
having heart disease (p = 0.01) were all significant predictors of continued
smoking at six months (Table 3). Factors that did not significantly predict
whether respondents were smoking at six months included: previous use of NRT (p
= 0.77); sex (p = 0.08, male:female, OR 1.49, 95% CI 0.92–2.42); ethnicity
(p = 0.12); who delivered the intervention (p = 0.51); years of smoking (p =
0.50); cigarette dose (p = 0.49); NZDep96 score (p = 0.08); reason for quitting
(range of p values 0.21–0.98); history of asthma or chronic obstructive
pulmonary disorder (p = 0.46); vascular disease (p = 0.31); and other
smoking-related disease (p = 0.46). The NZDep96 scores were collapsed into three
categories to test for any evidence of a trend: ‘low’ (NZDep96 score
1–3), ‘medium’ (NZDep96 score 4–7) and
‘high’ (NZDep96 score 8–10). There was no significant linear
trend (OR: low 1.00 (reference), medium 0.69, high 1.38;
χ2
= 1.14; p = 0.29).
Table 3. Univariate analysis
As seen in Table 4, subsequent multivariate analysis using a
logistic regression model showed there was interaction between variables that
seemed to predict outcome using univariate analysis. When applying the
multivariate model, the only independently significant predictors of smoking
status at six months were two with a negative influence on outcome. Having
others living in the house who smoked (OR 0.55, 95% CI 0.33–0.93, p =
0.03) and having made no previous quit attempts (OR 0.29, 95% CI
0.12–0.71, p = 0.02) were factors significantly associated with continuing
to smoke. Where respondents had made previous quit attempts, outcomes did not
significantly differ between different numbers of quit attempts.
Table 4. Multivariate analysis
The reason for restarting smoking most commonly reported was
stress (36%), followed by addiction to or satisfaction from nicotine/smoking
(20%), weight concerns (10%), and loss of motivation (10%). Withdrawal symptoms
were given as a reason for restarting by 6% of respondents.
When asked for suggestions for other support the IPA or
practice could provide that might make quitting easier, the most common response
was ‘none’ (35%), with 13% suggesting further follow-up
appointments, 9% suggesting more counselling, and 12% suggesting support groups
might be helpful. These responses were similar for both smokers and non-smokers,
except for the suggestion of support groups: nearly 90% of respondents who made
this suggestion were smokers.
DiscussionPreventive care is regarded as a
core feature of primary care and this study indicates that smoking cessation can
be very effectively delivered in the primary care context. The programme had a
good uptake, with enrolment of an average of 16 patients for every GP per year
(full or part time).
 Figure 1. Socioeconomic distribution of programme
participants
Figure 1 shows the socioeconomic spread of those enrolled in
the programme. It is reassuring to see that the programme is reaching the lower
socioeconomic group as it is not uncommon for middle and higher socioeconomic
groups to have a greater rate of uptake of preventive programmes. There is still
room for improvement, however, as proportions in this study sample still do not
match the nationwide proportions of smokers in the more deprived groups (Figure
2). The penetration of the programme into non-European ethnic groups is not so
good. Numbers in non-European ethnic groups were too small to allow
sub-analysis. In particular, Maori and Pacific Island enrolment rates were lower
than population proportions and Asian participants were notable by their
absence. This is of concern as Census data indicate that Maori have higher rates
of smoking in the New Zealand population than Europeans. The reasons for this
low uptake need further exploration.
Figure 2. Percentage of regular smokers in New Zealand
by NZ index of deprivation (source: Dr P Crampton, Health Services Research
Centre, Victoria University, Wellington)
 In this study, the population of participants was highly
mobile. Whilst this is not an unexpected finding it is important in planning
service delivery and programme follow up. It is interesting to note that while
deprivation is a predictor of starting smoking it did not appear to influence
programme outcome using NZDep96 as an indicator.
This programme compares favourably with six-month quit rates
for NRT-based programmes reported in the international literature of
14–22%. Rates in primary care settings are often at the lower end of the
range. Deception rates in self-report were tested in the previous study with
carbon monoxide assessments10 and were found to
be 14%, which is consistent with rates in the
literature3 and does not substantially affect
this judgement of effectiveness.
The reasons for the apparent success of the PEGS programme
cannot be inferred directly from this study, but some features of the programme
would seem likely to increase its potential for effect. The programme was
introduced as a modification of a well-established programme (the Smokescreen
Programme) after evaluation in a local
context.10 Key features of the modified
programme were: the flexible team approach, NRT subsidisation, and an effective
programme coordinator.
Use of the Di Clemente and Prochaska states of change model
ensures the programme is delivered to those who are more likely to be ready to
quit.9 A team approach gives flexibility of
role in delivery, ensuring that the most suitable person in the practice team is
able to deliver the programme. Increased PN involvement also reduces access
barriers by reducing costs and increasing potential contact time for the
patient. It would appear that programme delivery by PNs could be more cost
effective; however, care is needed in interpreting the comparative outcome
result. Practices in the study were given the flexibility to implement the
programme as it best suited their teams’ strengths and weaknesses and
prescribed roles would reduce this flexibility and could affect
outcomes.
The context within which the programme was developed was
also favourable; at the time, the ‘smoke free’ theme had a high
profile in NZ, with gradually increasing public acceptance of and demands for
more smoke-free areas, and with political support in legislation for smoke-free
environments. There had also been a significant increase in tobacco prices in
May 2000. The programme was branded as a locally modified product, and arrived
at a time when enthusiasm for the Pegasus Health IPA was high and there was good
practice-level acceptance. The application of an evidence-based approach to
practice was already well established and of proven effectiveness within the
organisation’s Clinical Practice Education Groups (CPEGs) to which all GPs
and most PNs belong. The programme coordinator provided training and resources
to practices and facilitated the implementation of the programme in a way that
was tailored to each particular practice’s working style and patient
population.
A response rate of 70% or greater is usually considered
ideal and, as with all studies, no conclusions can be drawn about outcomes for
those who were not contactable. As described, strenuous efforts were made to
contact all enrolled patients in the cohort, and it is clear that this is a
highly mobile patient group. Almost all the remaining non-responders had moved
and had no identifiable, current contact address.
Smoking cessation is very cost effective when compared with
other preventive interventions.14 This study
shows it can be effectively implemented in a primary care setting. The
effectiveness of an NRT-based smoking cessation programme in a general primary
care setting appears to have been significantly enhanced by local adaptation,
NRT subsidisation, use of the strengths of a flexible primary-care-team
approach, and effective coordination and facilitation responsive to individual
practice needs. The programme could be improved by testing the effectiveness of
the addition of the most common patient suggestions (longer follow up and
support groups).
A description of practices in primary care smoking cessation
in 2000 indicated a gap between reported and recommended practice in primary
care in this area.7 The success of this
programme in helping individual patients quit as well as its successful
implementation in a wide primary care setting suggests General Practice has an
important role to play in a country with a high burden of disease from
smoking-related illnesses. The PEGS programme is congruent with the 2002
Guidelines for Smoking Cessation endorsed by the Royal New Zealand College of
General Practitioners.15 Widespread adoption of
this kind of model in IPA/PHO settings throughout New Zealand should be
encouraged and, more importantly, supported. Pegasus Health is a large IPA that
has a well-developed infrastructure and resources that allowed it to develop and
facilitate the implementation of the PEGS programme. IPAs with different levels
of infrastructure development should be supported in providing the required
evidence-based education, coordination and practice communication necessary to
implement this type of programme.
Author information:
Dee A Richards, Senior Lecturer; Les J Toop, Pegasus Professor of General
Practice, Department of Public Health and General Practice, Christchurch School
of Medicine & Health Sciences; Keith Brockway, General Practitioner, Member
Pegasus Health IPA; Sue Graham, PEGS Smoking Cessation Programme Coordinator
Pegasus Health IPA; Bill McSweeney, General Practitioner, Director Pegasus
Health IPA; Donna MacLean, Practice Nurse, Convenor PEGS Committee Pegasus
Health IPA; Margaret G Sutherland, Research Coordinator; Alison Parsons,
Research Assistant, Department of Public Health and General Practice,
Christchurch School of Medicine & Health Sciences
Acknowledgements: We
thank the GPs and PNs of Pegasus Health who implemented this programme and the
patients for taking the time to complete the survey. The study was funded by
Pegasus Health IPA.
Correspondence: Dr
Dee Richards, Department of Public Health and General Practice Christchurch
School of Medicine & Health Sciences, PO Box 4345, Christchurch. Fax: (03)
364 3637; email: derelie.richards@chmeds.ac.nz
References:
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| | Current
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads) Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals |
|