Novel setting for addressing tobacco-related disparities: a survey of community welfare organization smoking policies, practices and attitudes.

Research in the United States and Australia acknowledges the potential of non-government social and community service organizations (SCSOs) for reaching socially disadvantaged smokers. This study aimed to describe SCSO smoking policies and practices, and attitudes of senior staff towards smoking and cessation. It also investigated factors associated with positive tobacco control attitudes. In 2009, a cross-sectional telephone survey was undertaken of senior staff in Australian SCSOs, 149 respondents representing 93 organizations completed the survey (response rate=65%; 93/142). Most service clients (60%) remained in programs for 6 months plus, and 77% attended at least weekly. Although 93% of respondents indicated they had an organizational smoking policy, it often did not include the provision of smoking cessation support. Most respondents indicated that client smoking status was not recorded on case notes (78%). Attitudes were mostly positive towards tobacco control in SCSOs, with a mean (standard deviation) score of 8.3 (2.9) of a possible 13. The practice of assessing clients' interest in quitting was the only statistically significant factor associated with high tobacco control attitude scores. The results suggest that SCSOs are appropriate settings for reaching socially disadvantaged smokers with cessation support. Although generally receptive to tobacco control, organizations require further support to integrate smoking cessation support into usual care. In particular, education, training and support for staff to enable them to help their clients quit smoking is important.

Introduction

Despite falling rates of smoking in most developed, industrialized countries, smoking remains one of the world’s leading preventable causes of illness and death [1, 2]. People with the highest smoking rates are those experiencing various forms of disadvantage [1-7]—for example, mental illness, poverty, indigenous or migrant heritage, imprisonment and substance abuse problems. Eliminating tobacco-related disparities has been identified as a public health priority in most Western nations including the United States, United Kingdom and Australia [8-10].

Reviews of smoking cessation treatment effectiveness have found it to be effective both for the general population (see Cochrane Tobacco Addiction Review Group) as well as selected disadvantaged groups such as low-income single mothers [11]. However, smokers experiencing disadvantage tend to have limited access to cessation support services and treatments [12]. For example, there is evidence of an ‘inverse care law’ whereby health services are more accessible in the more affluent (and often less needy) areas [13]. There is also evidence that patients from a lower socioeconomic position receive less preventive care from their doctor [14]. Similarly, fewer calls to national telephone cessation support services are made by people from a lower socioeconomic position [15]. The challenge is both to improve access to existing services and find new ways to reach smokers from disadvantaged groups. Murray et al. [16] systematically reviewed studies of strategies to improve access to smoking cessation support services for disadvantaged smokers and found the most successful strategy was the National Health Service-funded Stop Smoking Services [17] in the United Kingdom that concentrate services in deprived locations.

In many countries, UK style stop smoking services do not exist and alternative methods of supporting smoking cessation among smokers from disadvantaged groups need to be explored. Some programs in the United States and Australia have partnered with non-government social community service organizations (SCSOs) as a vehicle for reaching socially disadvantaged smokers [18-23]. Collectively, services for the homeless, people with severe mental illness and those with drug and alcohol problems, for example, have extensive reach into disadvantaged populations. Many provide holistic care and support and most have regular and often long-term contact with their clients [19]. Research has begun to show client and case-worker support for delivering smoking cessation care through this setting [20, 21] and one pilot study has demonstrated the feasibility and acceptability of SCSOs for smoking cessation [22].

It is generally unclear how receptive SCSOs would be to taking a more active role in addressing smoking among their clients and the degree of difficulty involved in translating their potential as sites for smoking cessation support into reality [23]. Organizational change theory suggests that the uptake of smoking cessation care provision by services requires endorsement by organizational leaders or champions; development of comprehensive organizational smoking policies and the introduction of systems to routinely assess, record and act on smoking status [24-27]. There is currently no information available regarding the smoking policies and practices within the non-government social and community services sector in Australia or elsewhere, or of the smoking-related attitudes of the service leaders. This study was undertaken with three aims: (i) to describe the smoking policies and practices within selected SCSOs in New South Wales (NSW), Australia; (ii) to describe the views of senior staff and management of SCSOs towards smoking and smoking cessation and (iii) to investigate factors associated with attitudes which are favourable towards smoking cessation support provision.

Methods

Design and setting

In 2009, a cross-sectional telephone survey was undertaken of SCSOs in NSW, the most populous state in Australia.

Sample

A list of organizations was obtained from the Council of Social Service of NSW (NCOSS). NCOSS is the peak body for the non-government social and community services sector in NSW [28]. The NCOSS database of 574 member organizations was screened according to the following eligibility criteria: (i) providing direct services to clients; (ii) having clients who were among the project target groups (disadvantaged youth, families, adults, indigenous Australians, mental health) and (iii) having a budget of at least AUD$300 000 per annum.

Screening yielded a sample of 142 eligible organizations stratified according to budget size—51 small organizations (budget of $300 000–$499 000); 56 medium organizations (budget of $500 000–$2 million) and 30 large organizations (annual budgets exceeding $2 million). Organizations with budgets of at least $300 000 were targeted for two reasons—they were more likely to be able to devote the time and resources to completing the survey and as a strategic and efficient approach by seeking to bring about change in the organizations that have contact with the most clients. Five smaller organizations (two with a budget of $50 000–$149 999 and three with a budget of $150 000–$299 000) were included at the request of NCOSS to represent remote services despite not meeting the budget size criterion.

Procedure

Initial contact was made with the executive officer of each eligible organization through a letter outlining the research and inviting the person to participate in a phone survey. A commercial company was commissioned to conduct telephone recruitment and interviews using computer-assisted telephone interviewing. All interviews were recorded. Participating executive officers from medium or large organizations were asked to nominate up to six managers from their programs to also participate in the research. The same information and consent process was used with the program managers and of those services. All participants were informed that the survey would take ∼20 min to complete and would involve questions primarily about the culture (attitudes), policy and practice regarding smoking in their organization. The study gained approval from the Cancer Council NSW Ethics Committee.

Measures

A telephone interview schedule was adapted from an instrument previously used in a similar setting [29, 30]. The instrument demonstrated good internal consistency (Crohnbach’s alpha coefficient = 0.80) and included items on the following:

Organization and manager characteristics: this included client target groups, number of paid staff and clients, types of programs and services, frequency and duration of client contact with the service [Program Managers (PMs) only], estimates of staff and client smoking rates and participant smoking status.

Organizational policies on smoking: eight items assessed whether the organization had a policy on smoking, and if so, what the policy included (no smoking indoors, in cars, with clients, on home visits, on outings and provision of cessation support) and where on the premises smoking was permitted.

Smoking cessation care provision to clients and staff: participants were asked to indicate whether staff in their organization offered smoking cessation support to clients from a list of eight interventions: recorded smoking status, assessed interest in quitting, provided cessation advice and counselling, emotional support and encouragement, participation in a quit group, referral to (telephone) Quitline, subsidized or free nicotine replacement therapy (NRT) and written materials. Participants were also asked whether the organization provides staff with smoking cessation training or resources to help them support their clients to quit smoking. In addition, participants were asked whether their staff were offered any form of support to quit their own smoking. If so, they were asked to identify the type of support: personal encouragement, access to (telephone) Quitline, written information, referral to (telephone) Quitline, free NRT and/or work-based quit counselling.

Attitudes and beliefs regarding smoking: participants were asked their position on 13 smoking-related statements (Table III) using a 5-point Likert-type scale (strongly agree, agree, neutral, disagree and strongly agree).

Table III.

Proportions ‘strongly agree/agree’ responses to smoking and tobacco control attitude items by organization size and total sample (n = 149 executive officers and program managers)

Attitude statement	Small (n = 42)		Medium (n = 74)		Large (n = 33)			Total (n = 149)
	n	% (95% CI)	n	% (95% CI)	n	% (95% CI)	P-value	n	% (95% CI)
(a) Disadvantaged people who smoke should receive help to quit	38	90 (81, 99)	66	89 (82, 96)	30	91 (81, 101)	NS	134	90 (85, 95)
(b) Because smoking is a personal choice, it is up to our clients whether they smoke or not, we should not interfere one way or another	21	51 (36, 66)	36	49 (38, 60)	10	30 (14, 46)	NS	67	45 (37, 53)
(c) Our clients who smoke are not really interested in quitting	14	34 (20, 49)	23	31 (20, 42)	5	15 (3, 27)	NS	42	28 (21, 35)
(d) Support to help people quit smoking should be a part of the normal care our organization provides	24	56 (41, 71)	52	70 (59, 81)	26	79 (65, 93)	0.044	101	68 (60, 76)
(e) Our clients are not really able to quit smoking	10	24 (11, 37)	13	18 (9, 27)	4	12 (1, 23)	NS	27	18 (12, 24)
(f) Smoking is something our program should give more attention to	23	54 (39, 69)	43	58 (47, 69)	18	55 (38, 72)	NS	83	56 (48, 64)
(g) Our staff have the confidence to provide smoking cessation advice and support to clients	27	63 (48, 78)	50	68 (57, 79)	20	61 (44, 78)	NS	97	65 (57, 73)
(h) Sometimes it is useful for a staff member to smoke with clients to build trust and rapport	4	10 (1, 19)	6	8 (2, 14)	4	12 (1, 23)	NS	15	10 (5, 15)
(i) Our staff have the knowledge and skills to provide smoking cessation advice and support to clients	17	41 (26, 56)	37	50 (39, 61)	18	55 (38, 72)	NS	73	49 (41, 57)
(j) For our clients, the benefits of smoking outweigh the disadvantages of smoking	9	22 (9, 35)	6	8 (2, 14)	5	15 (3, 27)	NS	21	14 (8, 20)
(k) Smoking increases our clients’ disadvantage	36	85 (74, 96)	56	76 (66, 86)	25	76 (61, 91)	NS	116	78 (71, 85)
(l) Smoking is not something our organization should give more attention to as we have other priorities	17	41 (26, 56)	25	34 (23, 45)	3	9 (1, 19)	0.005	45	30 (23, 37)
(m) Our staff have organizational support to provide smoking cessation advice and support to clients	21	51 (36, 66)	44	59 (48, 70)	22	67 (57, 83)	NS	88	59 (51, 67)

aSmoking cessation positive attitude was strongly agree/agree with statements a, d, f, g, i, k and m and strongly disagree/disagree with statements b, c, e, h, j and l.

Analysis

All analyses were conducted using SAS 9.0 statistical software. Executive officer and program manager responses were compared, and where no significant differences were found, were combined as averages.

Organization and manager characteristics, policy and practice outcomes

Categorical items are reported using proportions and 95% confidence intervals (95% CIs). Continuous items are reported using means and standard deviations (SDs). Differences across organizational size were assessed using generalized estimating equations.

Attitudes

Staff attitudes towards smoking and smoking cessation are reported for each of the 13 items using percentages and 95% CIs. For analysis, strongly agree and agree were collapsed into a single category ‘agree’ and strongly disagree and disagree were collapsed into a single category ‘disagree’ for each item. In addition to reporting of individual attitude items, an aggregated score [31, 32] was calculated to represent views that are supportive of smoking cessation provision or negative towards smoking. This was obtained by allocating a score of 1 to responses reflecting agreement with statements that were supportive of smoking cessation and 0 to those indicating a view that was less supportive (see Table III footnote). In this way, the maximum attitude score is 13. Means and SDs of scores are compared using a two-sample t-test, adjusted for clustering. Following examination of the distribution of scores, a cutoff for ‘high’ smoking cessation positive attitude score was set at 7 or more. Simple logistic regressions were conducted to examine factors that may be associated with high smoking cessation positive attitudes followed by a multivariate logistic regression model with those variables which resulted in significant outcomes. Factors that were examined using simple models were—do staff ask whether clients are interested in quitting? (almost always, often, sometimes and rarely); organization policy on smoking (yes and no); smoking in cars (yes and no); staff smoking with clients (yes and no); smoking on home visits (yes and no); smoking by clients on outings (yes and no); provision of quit support (yes and no); do staff record client smoking status (yes and no) and organizational size (large and small/medium). Only variables that had an unadjusted P-value of 0.2 or less were considered candidates for the stepwise regression. Odds ratios and 95% CIs were calculated for the predictor variables entered into the logistic regression. Goodness of fit of the logistic regression models was assessed using the Hosmer-and-Lemeshow test. All analyses were adjusted for clustering by organization.

Results

In total, 93 Executive Officers (EOs) and 56 PMs (that is 149 respondents) were recruited and completed the surveys, from the 142 eligible organizations. The flowchart represented in Fig. 1 provides a detailed illustration of the number of EOs and PMs recruited. Note that the 26 EOs recruited from ‘small’ organizations included those from the additional five remote/rural organizations.

Fig. 1.

Flowchart showing Executive Officer and Program Manager recruitment across different organizational sizes.

Organization programs, client and staff descriptions

Based on EO-only self-report (n = 93), the mean number of paid staff employed by the organizations was 79 (SD = 378, median = 13, range = 2–3600), the mean number of clients was 1778 (SD = 2944, median = 500, range = 18–15 000) and the total number of clients across 93 organizations was 148 983. PMs only (n = 56) were asked how long clients remained in their program and how frequently they visited the service. Most respondents indicated that clients remained in the program for 6 months or more (60%: 47%, 73%). During that time, most clients attended the program weekly or more often (77%: 66%, 88%). Some programs were drop-in centres or live-in programs (7%: 0, 14%).

The main client groups targeted by these programs were women (47%: 39%, 55%), children (47%: 39%, 55%), families (44%: 39%, 52%), homeless people (42%: 34%, 50%), vulnerable youth (41%: 33%, 49%), people with mental illness (41%: 33%, 49%), people with drug and alcohol problems (38%: 30%, 46%) and aboriginal people (28%: 21%, 35%). Comparisons of outcomes across target groups were not made due to the significant amount of overlap with services reporting to target two or more groups.

Participant smoking status and estimates of staff smoking status (N = 149)

Almost one in five (19%: 13%, 25%) respondents said they were daily or occasional smokers and a further 8% (4%, 12%) said they were ‘not regular/have tried smoking’ smokers, 46% (38%, 54%) said they were ex-smokers and 27% (20%, 34%) said they had never smoked. When asked to estimate the proportion of their staff they believed were current smokers, 19% (13%, 25%) of respondents indicated that no staff in their organization currently smoked, 50% (42%, 58%) of respondents estimated that up to 20% of their staff were current smokers and 27% (22%, 36%) estimated that over 21% of staff currently smoked.

Smoking-related policies (N = 149)

In total, 93% (89%, 97%) of respondents indicated that they had a smoking policy at their service. Table I shows the extent of smoking policies by size of organization.

Table I.

Proportions of respondents (n = 149 executive officers and program managers) indicating workplace smoking policies by organization size

	Small (n = 42)	Medium (n = 74)	Large (n = 33)		Total (n = 149)
	n	n	n		n
	% (95% CI)	% (95% CI)	% (95% CI)	P-value	% (95% CI)
Does the organization have a policy on smoking—yes	36	69	32	NS	137
	88 (78, 98)	93 (87, 99)	97 (91, 103)		92 (89, 97)
If yes, does this policy include the following:	Small (n = 36)	Medium (n = 69)	Large (n = 32)		Total (n = 137)
Smoking inside offices and buildings	36	69	32	N/A	137
	100 (96, 102)	100 (97, 101)	100 (96, 102)		100 (97, 101)
Smoking in cars	27	56	29	NS	112
	75 (62, 88)	81 (72, 90)	91 (81, 101)		82 (76, 88)
Staff smoking with clients	21	46	23	NS	90
	58 (43, 73)	67 (56, 78)	72 (56, 88)		66 (58, 74)
Smoking on home visits	18	39	24	0.05	81
	50 (35, 65)	57 (46, 68)	75 (60, 90)		59 (51, 67)
Smoking by clients on outings	13	27	8	NS	48
	36 (21, 51)	39 (28, 50)	25 (10, 40)		35 (27, 43)
Provision of quit support	14	25	15	NS	53
	39 (24, 54)	36 (25, 47)	47 (30, 64)		39 (31, 47)
Where in the service can clients smoke	Small (n = 42)	Medium (n = 74)	Large (n = 33)		Total (n = 149)
Anywhere inside and outside	0	1	0	NS	1
	0	1 (−1, 3)	0		1 (−1, 3)
Anywhere outside only	8	7	2	NS	17
	20 (8, 32)	10 (3, 17)	6 (−2, 14)		11 (6, 16)
Designated smoking areas	23	46	23	NS	92
	56 (41, 71)	62 (51, 73)	70 (54, 86)		62 (54, 70)
Off-site only—not on the premises	10	20	8	NS	38
	24 (11, 37)	27 (17, 37)	24 (9, 39)		26 (19, 33)

Provision of smoking cessation care to clients and staff (N = 149)

Most respondents said that client smoking status was not recorded on client case notes (79%: 71%, 85%). Significantly more large organizations recorded client smoking status (30%: 23%, 37%) than medium (19%: 13%, 25%) or small (17%: 11%, 23%) organizations (P < 0.05).

Just under half of all respondents (45%: 37%, 53%) indicated that staff in their organization ‘rarely’ ascertained clients’ interest in quitting. About a third (30%: 23%, 37%) said that staff ‘sometimes’ asked whether clients were interested in quitting and a quarter (25%: 18%, 32%) ‘often’ or ‘almost always’ asked clients about their interest in quitting. Table II summarizes the types of cessation support provided to clients who smoke.

Table II.

Types of cessation support provided to clients who smoke based on respondent self-report, by organization size (N = 149 participating executive officers and program managers)

	Organization
	n
	% (95% CI)
	Large (n = 33)			Medium (n = 74)			Small (n = 42)
	All/Many	Half/Some	None	All/Many	Half/Some	None	All/Many	Half/Some	None
Cessation advice and counselling	5	17	8	14	20	39	5	10	26
	16 (3, 29)	55 (38, 72)	26 (11, 41)	19 (10, 28)	27 (17, 37)	53 (42, 64)	12 (2, 22)	24 (11, 37)	63 (48, 78)
Emotional support and encouragement	10	17	4	26	27	20	12	19	10
	31 (15, 47)	53 (36, 70)	13 (1, 25)	35 (24, 46)	36 (25, 47)	27 (17, 37)	29 (15, 43)	46 (31, 61)	24 (11, 37)
Participation in a quit group	2	5	23	5	12	56	1	4	36
	7 (1, 16)	16 (3, 29)	74 (59, 89)	6.8 (1, 13)	16 (8, 24)	76 (66, 86)	2 (−2, 6)	10 (1, 19)	88 (78, 98)
Referral to quitline	10	16	5	18	40	15	11	21	9
	31 (15, 47)	50 (33, 67)	16 (3, 29)	24 (14, 34)	54 (43, 65)	20 (11, 29)	27 (13, 41)	51 (36, 66)	22 (9, 35)
Subsidized or free NRT	5	6	21	2	5	65	2	2	37
	15 (3, 27)	18 (5, 31)	64 (47, 81)	3 (1, 7)	7 (1, 13)	88 (81, 95)	5 (−2, 12)	5 (−2, 12)	90 (81, 99)
Written information	7	16	8	17	24	31	10	13	18
	22 (8, 36)	50 (33, 67)	25 (10, 40)	23 (13, 33)	33 (22, 44)	42 (31, 53)	24 (11, 37)	32 (18, 46)	44 (29, 59)

In most cases, respondents indicated that, in the 12 months prior to the survey, staff at their service did not receive organizational smoking cessation care training (80%: 74%, 86%) or resources (53%: 45%, 61%) to help them support their clients to address smoking.

Less than half of respondents (43%: 35%, 51%) indicated that staff were provided with any form of support to quit. When support was offered, the most common types were personal encouragement (81%), access to (telephone) Quitline (61%), written information (57%), referral to (telephone) Quitline (52%), free NRT (20%) and/or work-based quit counselling (14%).

Smoking-related attitudes (n = 149)

Table III shows the respondents’ views on a range of items relating to smoking cessation care provision, and smoking and disadvantage. In general, attitudes were mostly positive towards the provision of smoking cessation with a mean (SD) attitude score of 8.3 (2.9) with no statistically significant differences by organization size.

A backward stepwise logistic regression was conducted to examine factors associated with a high attitude score based on those that were significant at the 0.25 level from the bivariate models (Table IV). The practice of asking clients whether they are interested in quitting was the only statistically significant factor associated with high positive attitude scores. The odds ratio for high smoking cessation attitude was 4.16 (2.87, 5.46) among organizations that almost always ask about interest in quitting, 3.48 (2.48, 4.49) among those who often ask and 1.82 (0.84, 2.81) among those who sometimes ask compared with organizations that rarely ask clients about their interest in quitting.

Table IV.

Factors associated with high tobacco control positive attitude score (n = 149 executive officers and program managers)

		Simple models		Stepwise model
Predictor		Estimate (95% CI)	P-value	Estimate(95% CI)	P-value
Do staff ask if clients are interested in quitting	Almost always	4.17 (2.89, 5.45)	<0.001	4.16 (2.87, 5.46)	<0.001
	Often	3.51 (2.56, 4.45)	<0.001	3.48 (2.48, 4.49)	<0.001
	Sometimes	1.83 (0.83, 2.84)	0.0003	1.82 (0.84, 2.81)	0.0003
	Rarely	Referent		Referent
Staff smoking with clients	No	−0.06 (−1.8, −0.25)	0.0107
	Yes	Referent
Smoking on home visits	No	−1.29 (−2.34, −0.24)	0.0164
	Yes	Referent
Provision of quit support	No	−1.63 (−2.58, −0.69)	0.0007
	Yes	Referent
Do staff record client smoking status	No	−2.36 (−3.39, −1.33)	0.0001
	Yes	Referent

Discussion

To better understand the potential role that non-government SCSOs might play in providing smoking cessation aid to their clients, most of whom are disadvantaged, this study surveyed leaders of organizations in NSW, Australia, about their attitudes towards smoking and cessation support for their clients, and the current smoking policies and practices in their organization. A number of outcomes suggest that the SCSO setting is a potentially effective way to reach disadvantaged smokers. The study found that SCSOs have numerous (148 983 clients over 12 months across 93 organizations), frequent (77% of clients attending weekly or more frequently) and often continuing contact (60% of clients remaining in programs for 6 months) with clients. The study also found that attitudes towards the provision of smoking cessation care to clients were generally positive and that positive attitudes were associated with staff provision of smoking cessation care (asking whether clients are interested in quitting). The majority of organizations (93%) reported to have policies regarding smoking in the organization with restrictions on smoking inside service buildings. However, most respondents also indicated that the policy did not include a number of features important for its effectiveness at discouraging smoking such as specific guidance on staff smoking with clients and or the provision of quit support. The study found a few differences in attitudes and policies depending on the size of the organization. In terms of attitudes, respondents from larger organizations (79%) were more likely to agree with the statement that ‘Support to help people quit smoking should be part of the normal care our organisation provides’ than those from medium-sized (70%) and smaller (56%) organizations and less likely to agree with the statement ‘Smoking is not something our organisation should give more attention to as we have other priorities’ (9% versus 34% and 41%, respectively). Respondents from larger organizations (75%) were more likely to report having policies about smoking on home visits with clients than medium-sized organizations (57%) and smaller organizations (50%).

As no other survey of this type has been conducted in this setting, it is not possible to compare the current results directly with those of past studies. Similar surveys have, however, been conducted in similar settings. Most notably, staff surveys conducted in drug and alcohol treatment settings [29, 30] and mental health settings [33-35] have found similar attitude results. Generally, these studies reveal negative attitudes towards smoking and positive attitudes towards tobacco control efforts. Like our current results, some reservations are revealed in these studies as well such as beliefs that clients are not able to quit smoking and that smoking is a personal choice and it was not their place to interfere. The results suggest that, despite positive tobacco control attitudes in general, SCSOs continue to need education directed at countering those beliefs that have not yet changed.

Implications for policy and practice

The study suggests that SCSOs are a potentially important setting for reaching smokers who experience various forms of disadvantage. Tobacco control advocates should work with SCSOs to further develop comprehensive smoke-free policies. The development and enforcement of comprehensive organizational smoking policies is a key element for effective organizational change encouraging smoking care provision. There is evidence from the mental health service setting that services with clear and overt smoking restrictions that staff adhere to are more likely to provide comprehensive smoking care [36].

The study also revealed the need to turn positive attitudes into action—the delivery of smoking cessation care. The results showed that some smoking cessation care was occurring within SCSOs—the most commonly used strategies given to most clients included emotional support and encouragement, referral to (telephone) Quitline and the provision of written information. These are largely low-cost strategies, where resources can be obtained by SCSOs from health departments free of charge. Amplifying these activities could have a positive health impact at minimal additional cost. One key process in encouraging uptake of smoking care provision within an organization is implementing a system for routinely assessing and recording smoking status [24, 25, 36]. This survey found that most staff (79%) do not record client smoking status. If services are unaware of client smoking status, they are unable to identify those clients who require assistance in quitting. This was evident in the finding that only 25% of respondents said that clients who smoke are asked whether they are interested in quitting ‘almost always’ or ‘often’. Developing systems whereby SCSO’s routinely record smoking status at intake could have spill-over effects making it easier to identify smokers and routinely ask them whether they are interested in quitting. Computerized systems, supported through government funding, have shown promise in general practice [37] and general hospital settings [38] and should be considered in this setting. Health departments can also take a leading role in providing SCSO staff with training in the delivery of smoking cessation care at no cost to the service. In Australia, the government is providing some subsidized NRT through general practitioners. SCSOs can be made aware of this scheme to link clients with subsidized treatment.

These results provide preliminary data that larger organizations may be better equipped with the policies, systems and organizational culture to make changes towards tobacco control than smaller organizations. Organizational change research highlights the importance of organizational characteristics for successful diffusion of innovation [39, 40]. In the non-government social and community services context, larger organizations may have greater capacity to introduce smoking cessation care than smaller ones that may require additional assistance for implementing systems change.

Strengths and limitations of the study

Strengths of the study include its novel setting, state-wide sampling frame and relatively high consent rate. The main limitation with the study is its limited generalizability—it only surveyed managers from organizations that were relatively large (income of at least $300 000 per annum)—excluding very small services (with the exception of five small remote services) that make up the majority of organizations listed in the NCOSS membership list. Although community social service organizations in NSW are unlikely to be very different to those in other states, enhancing the transferability of these outcomes to other parts of Australia, it is unclear whether other countries have similar services and whether those services are equally prepared to address tobacco smoking. The field would benefit from cross-country comparison research. Also, managers and executive officers in some services were unreachable or refused to participate. This suggests that the current results may not be representing services that are ‘difficult-to-engage’. However, it is important from a public health perspective that the larger organizations that have the most access to high numbers of disadvantaged people who smoke and that are ready for change are targeted in order to produce the greatest effect with the least resources. Another source of potential bias is the study’s reliance on self-report—reports of smoking status are likely to be underestimates and reports of smoking cessation care provision are likely to be overestimates. Social desirability bias may also be influencing the favourable attitudes reported by participants and may not reflect the attitudes of the majority of employees who would deliver the smoking cessation services.

In conclusion, the results of this study were consistent with similar studies [18, 20–22] that show that SCSOs have potential as settings for reaching high numbers of smokers experiencing financial and social disadvantage. Moreover, the results show that some organizations may be receptive to integrating smoking cessation support and some have taken some steps in that direction. This is important given the social gradient in smoking prevalence rates and the need to consider new and innovative ways to reach smokers from disadvantaged groups. Further research is required to develop smoking cessation care systems that are suitable for implementation in the SCSO setting, and evaluation to assess the effectiveness of this setting at decreasing smoking rates among clients. Further research should also explore methods for linking SCSOs’ staff and clients with community-based smoking cessation services to improve access to cessation support.

Funding

Cancer Council New South Wales.

Conflict of interest statement

None declared.