Figures
Abstract
Financial (positive or negative) and non-financial incentives or rewards are increasingly used in attempts to influence health behaviours. While unintended consequences of incentive provision are discussed in the literature, evidence syntheses did not identify any primary research with the aim of investigating unintended consequences of incentive interventions for lifestyle behaviour change. Our objective was to investigate perceived positive and negative unintended consequences of incentive provision for a shortlist of seven promising incentive strategies for smoking cessation in pregnancy and breastfeeding. A multi-disciplinary, mixed-methods approach included involving two service-user mother and baby groups from disadvantaged areas with experience of the target behaviours as study co-investigators. Systematic reviews informed the shortlist of incentive strategies. Qualitative semi-structured interviews and a web-based survey of health professionals asked open questions on positive and negative consequences of incentives. The participants from three UK regions were a diverse sample with and without direct experience of incentive interventions: 88 pregnant women/recent mothers/partners/family members; 53 service providers; 24 experts/decision makers and interactive discussions with 63 conference attendees. Maternity and early years health professionals (n = 497) including doctors, midwives, health visitors, public health and related staff participated in the survey. Qualitative analysis identified ethical, political, cultural, social and psychological implications of incentive delivery at population and individual levels. Four key themes emerged: how incentives can address or create inequalities; enhance or diminish intrinsic motivation and wellbeing; have a positive or negative effect on relationships with others within personal networks or health providers; and can impact on health systems and resources by raising awareness and directing service delivery, but may be detrimental to other health care areas. Financial incentives are controversial and generated emotive and oppositional responses. The planning, design and delivery of future incentive interventions should evaluate unexpected consequences to inform the evidence for effectiveness, cost-effectiveness and future implementation.
Citation: Thomson G, Morgan H, Crossland N, Bauld L, Dykes F, Hoddinott P, et al. (2014) Unintended Consequences of Incentive Provision for Behaviour Change and Maintenance around Childbirth. PLoS ONE 9(10): e111322. https://doi.org/10.1371/journal.pone.0111322
Editor: Jacobus van Wouwe, TNO, Netherlands
Received: May 16, 2014; Accepted: August 19, 2014; Published: October 30, 2014
Copyright: © 2014 Thomson et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. The University of Aberdeen (study sponsors) and the HTA (study funders) actively encourage collaboration. Data excerpts are presented in the manuscript. The full dataset cannot be publicly available for ethical reasons: public availability would compromise participant privacy. Additional supplementary de-identified quote material to support the findings and conclusions of the paper are available on request. Interested researchers can make an application to the PIs for the study: Prof Pat Hoddinott, (University of Stirling, p.m.hoddinott@stir.ac.uk) and Dr Heather Morgan (University of Aberdeen, h.morgan@abdn.ac.uk) to discuss data sharing and their data requirements. External users will be bound by a University of Aberdeen data sharing agreement, which will be in place prior to release, and will be required to work with the research team to comply with the ethics committee approval for this study. A University of Aberdeen data sharing platform will be used to enable researchers who are given permission to access the data in a secure format.
Funding: This report was commissioned by the NIHR HTA Programme as project number 10/31/02. http://www.nets.nihr.ac.uk/projects/hta/103102. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
There has been growing international academic and policy interest in the use of financial incentives to change health behaviours [1]–[3]. Similarly, there is evidence, particularly from the UK Quality and Outcomes Framework, that financial incentives to care providers can change health professional behaviour resulting in improved health outcomes; however to date disease rather than ‘lifestyle’ behaviour outcomes have been the focus [4], [5]. Literature on the mechanisms of action of incentive interventions suggests that incentives can “crowd out” intrinsic motivation [6]–[8] when financial incentives are viewed as paternalistic or undermining autonomy when recipients are ‘told what to do’ [9].
‘Unintended consequences’ is a key area of contention within incentive delivery. The phrase was first coined by Merton [10] and refers to outcomes other than the ones intended by a purposeful action. The purposeful actions intended by financial incentives in health care are changes in behaviour (at citizen, patient, healthcare provider or organisation level) which lead to evidence of improved health outcomes (effectiveness and/or cost-effectiveness). However, negative consequences have been reported, particularly by the media which describe incentives as potentially coercive or encouraging unhealthy behaviours or game playing to ensure eligibility [11]–[13], generating public and tax payers’ debate about the appropriateness of incentives, particularly in countries with state-funded health services.
Systematic reviews conducted as part of the BIBS (Benefits of Incentives for Breastfeeding and Smoking Cessation) study, which aimed to inform the design of incentive intervention trials for smoking cessation in pregnancy and breastfeeding, provide the background to this paper and are reported in full elsewhere [14], [15]. Incentives were defined as ‘financial (positive or negative) and non-financial tangible incentives or rewards, such as free or reduced cost items or services that have a monetary or an exchange value’ [14]. Systematic reviews investigated i) the evidence for the effectiveness of incentive interventions delivered within or outside the health service, to: a) individuals, families or b) organisations that aim to increase and sustain smoking cessation and breastfeeding [14] and included a narrative review of qualitative and process evaluation data; ii) a narrative synthesis of qualitative reviews reporting barriers and facilitators to smoking cessation and breastfeeding; iii) a scoping narrative review of reviews of the effectiveness of financial incentives for other lifestyle behaviours relevant to women of childbearing age.
Data relating to the consequences of incentive interventions (intended and unintended) were extracted and analysed in these reviews. In summary, the evidence syntheses did not identify any primary research with the aim of investigating unintended consequences of incentive interventions for lifestyle behaviour change. Positive and negative unintended consequences of incentive provision were sometimes referred to within the discussion section of the papers. Gaming and cheating were reported as a concern for incentives to influence smoking cessation, particularly in high risk populations, (e.g. the vulnerable/those with chaotic life styles) [16]–[18]. They may be considered ‘unfair’ for those who make healthy choices [19], [20]. While financial incentives to health care providers are considered to narrow health inequalities between the most and least deprived populations [4], [21], a review of the evidence for incentive schemes to encourage positive health and social behaviours in young people identified that incentives targeted some groups and not others and this was perceived as unethical and inequitable [22]. This review [22] and a smoking cessation in pregnancy incentive intervention delivered through pharmacies [23] suggest that incentive programmes may not achieve their intended demographic reach, with more advantaged groups benefiting from the programme. Furthermore, if behaviour change is not achieved, self-esteem may be reduced and incentives that are negatively perceived can cause harm, such as undesirable peer pressure and bullying [22], [24].
Incentives to providers have been reported to inadvertently promote unethical ‘gaming’ behaviour, through distortion, manipulation or concealment of data [25], [26]. While incentives can improve the performance and cost-effectiveness of care on targeted conditions [4], large scale/complex incentive interventions are administratively labour and resource-intensive [22], [27]. Concerns have also been raised about the neglect of non-incentivised conditions or patients for whom the conditional quality target for providers to receive the incentive is more difficult to achieve [4], [26]. Evidence suggests that care provider efforts tend to wane once the target has been achieved [4] and that improvements occur at the fastest rate in the first year of a programme and subsequently return to the pre-intervention rates of improvement [4], [21]. Incentives can improve teamwork, enhance specialist skills [4], [18], [28] and facilitate connections between providers and consumers [29]. However other studies report reductions in person-centeredness, patient satisfaction and continuity of care [4], [30] as well as expectations for financial reimbursement across other areas of health care [26].
In this paper, we report on mixed methods primary research on the positive and negative unintended consequences of financial incentive provision (to consumers and providers) where the intended consequences are defined as smoking cessation and breastfeeding behaviour change and maintenance around childbirth and the unintended consequences concern outcomes other than those intended. ‘Consumers’ refer to the target population whose behaviour the incentive aims to change. Some, but not all, will be health service patients. ‘Providers’ refer to any staff, services or organisations which have a role in supporting women to stop smoking in pregnancy and/or to breastfeed after birth.
Materials and Methods
Ethics
Full ethical approval for this study, including service user involvement, was obtained from the North of Scotland Research Ethics Committee (NOSRES, reference number: 12/NS/0041, 12th April 2012), and subsequent permissions were granted locally by Research and Development, NHS Grampian (24th April 2012) and the BUSH (Built & Natural Environment, Sport and Health) Ethics Committee, University of Central Lancashire (BUSH064, 8th May 2012). Amendments were submitted to NOSRES. AM01 to cover the amendments required by BUSH (approved 10th May 2012); AM02 to allow us to use a flyer for recruiting health professionals at conferences and an information leaflet designed for partners/family/friends (approved 6th December 2012) and AM03 to gain ethical approval for the contents of the general public and health professionals survey (approved 17th April 2013). Participants who took part in the interviews, focus groups or interactive discussions provided written (for face to face interviews/focus groups) or verbal (for telephone interviews, interactive discussions) informed consent to participate in this study. Survey respondents were not requested to complete a consent form, rather consent was implied by their participation. All consent procedures received ethics approval.
Study design
Three evidence syntheses described above were integrated with primary qualitative and survey research to investigate diverse perspectives on incentives for smoking cessation in pregnancy and breastfeeding using a multi-disciplinary, mixed methods approach. Researchers regularly engaged two mother-and-baby groups in disadvantaged areas (study co-applicants) who provided diverse, hard-to-reach service-user involvement.
Qualitative data
Setting and participants.
Three settings (Lancashire, Glasgow and Aberdeen) were purposively selected for their diverse socio-demographic characteristics and their different incentive cultures for smoking cessation in pregnancy and breastfeeding (Table 1 provides an overview of the demographics and characteristics of the three selected sites). A third sector incentive scheme designed for young parents had been implemented in Aberdeen. Health service incentive schemes for both target behaviours had been operating in Lancashire. In Glasgow, a concurrent health service smoking cessation Phase II incentive trial was being undertaken with a qualitative process evaluation [34].
Qualitative research involved purposive, theoretical and snowball sampling undertaken by five interviewers (GT, NC, HM, JMcK, SM) across the three sites to include individuals with and without experience of incentive interventions. Participant recruitment was facilitated by discussions with service managers and key workers in health, social and third sector organisations. There were three sample populations: a) pregnant women, new mothers and their partners/family members who had become parents within the last 6 months; b) providers who could either deliver or receive incentives to support women to initiate or maintain smoking cessation or breastfeeding and c) experts and decision makers either in a management/coordinator position potentially responsible for implementing incentive programmes, e.g. a policy maker at local, regional or national level or a member of a research ethics committee.
Data collection.
A range of qualitative methods were used between November, 2012 and June, 2013, including unstructured semi-structured and structured interviews with vignettes, focus groups, interactive discussions and four open questions on a web survey [14]. Our iterative approach to data collection and analysis continually generated new research questions. For each question we discussed as a research team the most appropriate qualitative method for data collection. For example, to understand unintended consequences for specific promising interventions identified in the evidence synthesis, we used vignettes to describe the aspects we were interested in and identified participants (i.e. women, partners, providers) who potentially could have been involved in such an intervention to ascertain their views. To seek disconfirming data, we conducted more unstructured individual interviews, rather than a group discussion where there is a tendency towards consensus, i.e. FG11 which involved a focus group with health visitors from one geographical location.
Semi-structured interviews were used to explore participants’ knowledge, experiences and attitudes towards incentive provision and the potential implications and consequences of incentives. Topic guides were designed with the involvement of service users and were modified iteratively as the study progressed. Eight vignettes were developed from studies identified in the systematic reviews, which were selected either because they had statistically significant effects or involved an unusual, promising or innovative approach. Six vignettes were used with mothers/partners/family members [37]–[42] and four with professionals [34], [37], [39], [43] (Table 2). The vignettes were included in interviews and focus groups to garner participants’ perspectives of the incentive intervention. On two occasions, interviews were held in the women’s homes; the remaining interviews/focus groups were held at community or health and social care services locations (e.g. mother and baby group, health clinic) or via the telephone. Interviews/focus groups ranged from ∼15 to 100 minutes and were audio recorded and transcribed in full.
GT, HM, NC and PH facilitated and recorded interactive discussions at three conferences (maternity/early years and public health) comprising academic, third sector and health professionals over the course of the study period to discuss the feasibility and acceptability of incentive provision for consumers and providers.
Survey data
Setting, participants and data collection.
The web survey was distributed to maternity and early years staff working in Scotland and North West England (see Appendix S1 for a full copy of the survey). Recipients were identified and contacted by health service gatekeepers to email lists for maternity services, child health and primary care via NHS Research and Development Networks, the Scottish Primary Care Research Network (SPCRN) and a private company. The emails provided an introduction to the study with a link to Survey Monkey (www.surveymonkey.com). With the exception of those reached through SPCRN, to whom one reminder was sent by each of the regional co-ordinators to their respective area email lists, no repeat emails were sent. All respondents were offered entry into a draw to win one of forty £5 retail vouchers.
Survey design.
A shortlist of seven promising incentive strategies emerged from the BIBS study evidence syntheses [14], [15], service-user feedback and early qualitative data collection. This shortlist informed the incentive descriptions used in the survey to investigate health care professional acceptability and anticipated consequences [14]. Survey questions asked participants to respond using a 5 point Likert scale (1- strongly agree; 5- strongly disagree) to seven promising incentive strategies identified from evidence syntheses and qualitative interview data. Five were incentives to women, including a free breast pump and conditional shopping vouchers for verified proof of smoking cessation at different time points and having a smoke-free home. Two were incentives to providers: (i) payments to local health services for reaching smoking cessation in pregnancy targets and (ii) breastfeeding targets. Demographic data and experience of target behaviours were recorded. To minimise framing effects, free text questions about positive and negative consequences were included, rather than a single question on unintended consequences. The free text survey questions were: We would like you to imagine that your local health service is going to run a scheme that provides incentives for stopping smoking in pregnancy. What do you think the consequence might be for participants and/or staff? Qi) Positive consequences? Qii) Negative consequences? Identical questions with breastfeeding inserted instead of stopping smoking in pregnancy followed.
Data analysis
The first stage of analysis used a Framework approach [44] to interpret the experiences and views of participants. A key strength of the framework approach is its potential to allow data to be summarised within thematic matrices and for patterns or explanations to be identified. All qualitative interview data were entered into NVivo10 software (QSR International, Burlington, MA) to facilitate data organisation, coding and retrieval. Free-text responses to open questions in the health professional survey on the perceived consequences of incentive programmes were entered onto an Excel chart and content analysis was used to triangulate the analysis of the interview data. Data cleaning was undertaken on the survey responses in that all comments not relating to the question were recorded as ‘unsure’ to calculate a response rate for each incentive consequences question. Initially, researchers (GT, NC, HM) identified key themes and categories independently by listening to and reading transcripts of the first four participant and four provider interviews. Through wider research team transcript reading and discussion, a single tree structure coding index was agreed and applied in NVivo10 to the separate site datasets, with 2–4 weekly merges of datasets [14]. The researchers undertook a detailed analysis of data with regular discussion several times a week between sites to ensure consistency and to search for disconfirming perspectives. Drafts of the findings and analysis were circulated prior to weekly meetings with feedback provided by the project lead (PH). Analysis continued until data saturation was achieved.
A second stage of more in-depth data analysis was undertaken for this investigation of the consequences of incentives. Braun & Clark’s [45] thematic analysis was undertaken which involved reading and re-reading of all the transcripts and free text responses, followed by coding, organising and mapping the data into groups and networks until saturation occurred. Initial data analysis was undertaken by GT and the findings were shared and discussed with NC, HM and PH for consensual validation of the final emergent themes.
The qualitative research was conducted or overseen by social science and/or health researchers, three of whom had been involved in incentive interventions (GT, LB, and PH). The research team included previous smokers, those with and without children, experiences of breast and formula milk feeding who held different perspectives on incentive interventions for behaviour change. Differences and potential biases were discussed in regular team meetings and noted in reflective diaries kept by the qualitative research team.
Results
Qualitative participants
A total of 177 participants took part in 16 focus groups, 55 face-to-face interviews and 19 telephone interviews (Table 3). This sample included 88 pregnant women/recent mothers/partners/wider family members; 53 service providers, 24 experts/decision makers and approximately 63 conference attendees participated in an audio-recorded interactive discussion. Our interview participants represented women living in disadvantaged areas recruited within and outwith health services with experience of smoking in pregnancy and choosing not to breastfeed, ethnic diversity and educational level (Table 4). Thirty women/parents had experience of an incentive intervention (22 in the CPIT trial/incentive intervention for smoking cessation; four in a NW England breastfeeding incentive scheme; four in a third sector teenage mothers’ programme). Twenty-three CPIT providers/experts participated. Two experts had been involved in a different voucher incentive programme for smoking cessation and one expert was involved in a breastfeeding incentive intervention. More detail linking the sample characteristics to ID codes and the characteristics of who took part in the interviews and focus groups are presented in Tables 5–8.
Survey participants
A total of 497 health and early years professionals responded to the survey. The characteristics of the respondents are reported in Table 9. The response rates to the free text questions are detailed in Table 10. This table indicates that survey participants were more likely to record positive rather than negative consequences for both target behaviours. Most positive comments concerned the expected health benefits associated with not smoking and breastfeeding, and the intended consequences of incentive delivery of smoking quit rates and increased breastfeeding.
The four emergent themes that integrated positive and negative unintended consequences are: addressing or creating inequalities; enhancing or diminishing motivation and wellbeing; relationships with others; and impact on health systems and resources. These are summarised in Table 11. In the following sections, each theme is described and illustrated with quotes from participants. Participant quotes have been assigned a code, for example (FG5, I, mother), which describes whether the participant took part in a focus group (FG), interactive discussion (IA), survey (S), telephone interview (T), or face-to-face interview (no code), and gives the participant identification number. The code also indicates whether or not the participant had been involved in an incentive programme (I versus no letter) and describes participant characteristics: professional background, whether the mother was pregnant for the first time (pregnant mother), a mother (who may or may not have been pregnant again) or a wider family member. Where appropriate, the findings have been integrated across the different participant groups. Where distinctions between participant groups are relevant, the term ‘consumer’ has been used for ‘women/partners/family member’ comments, and ‘professional’ relates to comments from providers/experts/conference attendees.
Addressing or Creating Inequalities
Financial incentives for smoking cessation and breastfeeding when discussed in relation to the Tappin et al. [34] and Heil et al. [39] intervention vignettes were considered to have ‘appeal’ particularly amongst teenage mothers and/or families ‘where money is an issue’ to ‘attract’ or ‘persuade’ them to engage in health services and for a ‘meaningful conversation’ to enable them to make ‘informed choices’ regarding their health behaviours:
I think if you were young, or if you were on your own and you might feel a big judged at times, or a bit….you know, why’s this person coming to look at me again, you know and I think the incentive scheme can only kind of help that really and make it a nicer experience (T9, I, mother).
Professionals perceived consumer incentives to be a ‘foot in the door’, leading to a ‘greater uptake of services’ particularly for ‘those in need’, leading to ‘improved health outcomes and opportunities to engage in other health advice’. However, the capacity to increase health inequalities due to marginalised families and those with very chaotic lifestyles being less likely to be aware of, and engage with, incentive provision was a concern amongst a number of the professionals:
It’s those that understand the system that benefit most and they’re the ones that least need it. So if there was an incentive scheme you can be pretty sure that everyone earning between £25,000 and £70,000 a year will be taking advantage of that incentive scheme whereas those who are on £8,000 a year won’t even know about the incentive scheme (FG12, providers & expert).
Concerns were raised by participants that incentive withdrawal consequent on continued smoking or relapse, or on reduced or discontinued breastfeeding could lead to women being less likely to ‘report problems’ or ‘non-engagement’ due to professionals ‘going on at me to give up [smoking]’ or continue breastfeeding.
Incentives for smoking cessation were considered by many of the participants to be ‘unfair’, for example, by ‘rewarding’ smokers who are doing ‘something they know to be detrimental to health’ or penalising those who ‘are doing the right thing’. Some perceived this as counter to parenting and education practices in terms of ‘positive reinforcements’ for good, rather than negative behaviour:
If you are a pregnant mum and not smoking, you should be incentivised because you are being the role [model] so that might give the others mums who smoke motivation to stop, knowing what they could get so you are rewarding the good behaviour (FG2, mothers).
Many professionals also raised concerns when discussing the Cattaneo et al intervention vignette [43] about how incentives to providers associated with meeting targets for smoking cessation and breastfeeding would be ‘inequitable’, ‘unrealistic’ and ‘unfair’ when women’s decisions and choices were considered largely outside of their control, particularly for those working in areas of high deprivation where smoking and formula feeding are more prevalent.
Other participants felt that incentives could ‘stigmatise’ and create ‘polarisation’ and ‘discrimination’ between different groups of women (e.g. those who breastfeed and those who formula feed), and even ‘resentment’ if the incentive was targeted towards a particular, ‘undeserving’ population: ‘well that’s my money, going to my next door neighbour, they don’t have a job’.
For geographically targeted incentives, some participants expressed concerns towards a ‘postcode lottery’ of care and believed ‘equity’ to be important as everyone needs support. However, universal incentives could ‘benefit those who already had enough’. Some professionals believed that targeted provision could help address embedded ‘social norms’ associated with the target behaviours. Many participants considered how incentives had the potential to reduce inequalities through providing access to items that they could not afford (e.g. breast pump, nursing bras), and financial support for those who are ‘struggling for money’ to buy essentials such as ‘food’, ‘things for the baby’ or for a ‘healthier lifestyle’:
Obviously it is an expensive thing, having a baby and I think that people who maybe need the income more, it would help massively. If I’d been in that boat it would have helped massively (44, I, pregnant woman).
Enhancing or Diminishing Intrinsic Motivation and Wellbeing
Numerous participants, in discussion of intervention vignettes such as Tappin et al. [34] and Heil et al. [39] believed that incentives for smoking cessation and breastfeeding could operate as a ‘tipping point’ to ‘encourage’ and provide an ‘extra boost’ for women to adopt healthy behaviours; with the ongoing delivery of incentives providing ‘something to look forward to’ to ‘push you [woman] a bit further’. Incentives could provide recognition of providers and consumers’ achievements promoting ‘well-being’, ‘self-esteem’ for women, and for providers ‘job satisfaction if the rates improve’.
Some consumers and providers who had been involved in incentive schemes for smoking cessation and breastfeeding referred to how women felt ‘privileged’, ‘valued’ and more ‘confident’:
I felt quite privileged they picked me for any group [research study incentive group] […] I felt quite happy that they had actually considered putting me in one of the groups because I never thought I was going to get into any of the groups to start with (43, I, pregnant woman).
A few professionals reflected that incentives would provide vulnerable individuals with one of the first opportunities to receive a reward and acknowledgement for an achievement. Unrestrictive incentives like shopping vouchers could provide the most disadvantaged families with a rare opportunity for autonomy to ‘make decisions about what they ought to be spending the additional money on’, such as providing ‘treats’ for themselves and their families.
Other professionals when discussing the Cattaneo et al vignette [43] considered how provider incentives could diminish health professional’s intrinsic motivation to support the target behaviours:
Then they [health professionals] are not actually being motivated by increasing people to breastfeed, they are going to be motivated by the fear of the humiliation if they don’t get there (IA2, providers & experts).
A number of the participants believed that incentives for smoking cessation could diminish women’s personal motivation by discouraging quitting prior to enrolling in the programme, thereby reducing ‘health choices to a financial transaction’, or even incentivise ‘people to get pregnant’. While some consumers considered this unlikely; ‘why damage your body if you don’t actually already do it’, incentives of higher value were generally believed to be associated with a greater likelihood of ‘gaming and cheating’ ‘but £750, like I say I would I start smoking for that’. Relapse after the incentive had been withdrawn, due to ‘people stopping for the wrong reasons’ and re-access ‘get the voucher, spend it, then start again’ were highlighted by many.
Participants also raised apprehensions with regard to vouchers that had a ‘currency’ value and could be exchanged for inappropriate items, e.g. ‘cigarettes’, ‘formula milk’ ‘illicit drugs’ or ‘alcohol’. Restrictions on voucher use, such as within the CPIT trial where general shopping vouchers could not be used for cigarettes or alcohol, or behaviour related incentives (i.e. a breast pump discussed in the context of the Chamberlain et al [37] intervention vignette) or small low value personal gifts were felt to be less open to abuse:
You can imagine they are probably open to some sort of manipulation especially if the rewards are financial rather than the recognition of a mile stone, a badge, a picture frame (T6, I, mother).
Other ‘gaming and cheating’ concerns were raised by participants with regard to how provider incentives may lead to staff ‘skewing’ their records, or women taking ‘advantage’ of the schemes, through making ‘fraudulent claims’ and ‘lying’ to ensure eligibility. These anxieties were magnified due to the perceived fallibility of ‘proof’; ‘no real way of knowing if women are actually continuing to breastfeed’. For smoking, there was some concern expressed that women may learn how to cheat when given Carbon Monoxide (CO) breath tests to confirm smoking status. CO can only capture recent smoking (in the past 12 hours) and there was some speculation that women might be able to abstain just for that period to obtain the incentive. However, this issue was not evident within the CPIT trial data from service providers:
So far the people that I have engaged with that have signed up to the service to me, bar one, I feel have been one hundred percent genuine (T72, I, smoking cessation advisor).
Many participants considered that verification of behaviour change was crucial to prevent the schemes coming into ‘disrepute’.
Relationships with Others
When discussing the Walsh et al. [42] and Cattaneo et al. [43] intervention vignettes, some participants considered how incentives to providers to meet targets for smoking cessation in pregnancy and breastfeeding could contribute to a ‘shared aim’ across different individuals and services and getting ‘everybody on board in some shape or part’. Incentives to consumers could also operate as ‘enablers’ for ongoing contact with services:
Sometimes, when you are feeling like rubbish and your house is a mess and you think “oh I can’t be bothered with somebody else coming round now”, you’ve got the health visitor and you’ve got you know all the other…. and you think “oh another person coming round to look at me”, but I think the incentives definitely can well …I will get a present from this one (T9, I, mother).
Regular women–provider contact, such as depicted within the Heil et al. [39] intervention vignette and the opportunity to demonstrate ‘value’ to women through incentive delivery was believed to help create a ‘positive effect’ on the women–provider relationship, assist women-centred care and make it ‘easier for staff’ to encourage women to adopt healthy behaviours.
From a counter perspective, women ‘being paid to do something that they should do without question for the health of the baby’ could cause resentment amongst providers. A few professionals voiced concerns of feeling ‘embarrassed’ in ‘selling’ incentives to women. Many consumers as well as professionals in response to the Cattaneo et al intervention vignette [43] reported that provider incentives could lead to women ‘feeling bombarded’, ‘bullied’ or ‘inappropriately handled’ in attempts to ‘manipulate people into a particular behaviour’; which in turn could exacerbate unhealthy behaviours, i.e. women ‘smoking even more’ or being less inclined to breastfeed. Other concerns were how providers may incorporate ‘bias and opinions rather than research and fact’ and provider–women interactions relegated to a ‘tick-box’ exercise; with efforts focused on targets ‘rather than healthcare’ and neglecting non-target areas: ‘bottle-feeding women having even less support’.
Some participants considered how incentives delivery could negatively impact on the ‘therapeutic relationship’ creating ‘distrust’ between women and providers, with women left feeling ‘judged and alienated from their health care givers’:
You’ve got to think about how these mums feel when they don’t want to breastfeed at all, then you’re bombarding them with it, and they won’t have any relationship and you’ve lost every other good thing you could do with them (FG11, health visitors).
A few participants reported that women’s involvement in incentive schemes could create ‘social connectivity’ through being ‘part of something’ and knowing that their involvement ‘could help other people’. However, failure or relapse could create additional strain on women through fears of ‘letting her [provider] down’. Withdrawal or non-eligibility of incentive schemes for women who were unable to breastfeed or quit smoking was considered by many participants to ‘penalise those who have tried’ creating ‘stress’, ‘pressure’ and ‘blame’, leading to ‘reduced self-esteem’, ‘depression’ ‘guilt’ and feelings of ‘failure’ and ‘inadequacy’:
Breastfeeding is not always a personal choice whether you can or not, whereas smoking you are personally responsible, whereas breastfeeding no matter how hard you try you might not be able to do it (17, grandmother).
A number of professionals, in reflection of the Cattaneo et al intervention vignette [43] believed that provider incentives would create stress through staff having to ‘confront’ and ‘challenge’ those whose ‘claims are disproved’. Some professionals also expressed that the pressure of target attainment could ‘demoralise’ staff, leading to a situation where they ‘disengage and don’t deliver effectively’ or even result in an ‘increase in sickness levels’.
Many participants considered incentive provision to be ‘patronising’ and to ‘infantilise women’ through re-enforcing a ‘paternalistic’, ‘nanny state’ ethos. The potential for creating new social or health care norms could potentially diminish individual responsibility for making healthy choices and shift the balance ‘from being a responsible mother to “the NHS will sort me out”’:
We’re health, you know. It’s your own health, take some responsibility. I think we’re going right down the wrong route; I think that, you know, when we’re enticing people with money and gifts just to do what’s right for their health’ (53, midwife).
Signing up to an incentive scheme was believed to have wider benefits for healthy behaviours ‘cascading’ to staff, family members and other members of the women’s personal networks; creating ‘health improvement for all the family’ and ‘healthy competition’ in achieving behaviour change. Consumers and professionals through discussion of the Gulliver et al. [38] and Heil et al. [39] intervention vignettes for smoking cessation also felt that while it was harder for partners to stop smoking ‘as a woman has a reason to’; the potential to engage them could help to share the responsibility for health, rather than the sole onus being placed on the woman:
Both of them are in it together then, they’re not feeling as if one is responsible for this if the baby is compromised, you know, I would hope even if they didn’t stop smoking, which I hope that they would, but if the baby was compromised they couldn’t blame one or other. It’s meeting the point that both of them are responsible for this baby (T59, midwife).
Participants also expressed ‘vested interests’ in supporting incentive schemes to prevent passive smoke exposure ‘being surrounded by people puffing away on fags’; encouraging women to ‘set a good example’, ‘raise community awareness’ and create ‘positive cultural change’ by normalising not smoking and breastfeeding:
If breastfeeding becomes more, people start thinking of breastfeeding once again as the norm then it will be easier for other people to do it, until it gets to such a point where it will be considered unusual to be going onto formula feed a baby straight after birth (T6, I, mother).
Conversely, some participants felt that the adoption of positive health behaviours could lead to social isolation for women ‘if all their family and friends are smokers’. The women’s efforts may be thwarted by ‘unsupportive partners and family members’ and have adverse consequences for personal network relationships with ‘everyone smoking around me’. The health benefits of quitting smoking in pregnancy were seen by some as subsequently ‘erased’ if the baby still resides within a ‘smoking household’.
A few professionals raised concerns about ‘domestic abuse’ in terms of how women may be pressurised by partners to sign up when they are ‘maybe not really 100% wanting to’ or for ‘people taking those vouchers off those women as they walk out the door’.
Impact on Health Systems and Resources
Investment in incentive programmes, despite the ‘hard strapped’ situation of the UK NHS was perceived by many of the participants to be an important endorsement of the ‘seriousness’ of not smoking and breastfeeding and a positive ‘pro-active’ rather than reactive stance to health promotion. A number of the participants considered how incentives were not ‘a lot of money’ compared to the long-term savings on health-related conditions associated with smoking and not breastfeeding; although some felt that evidence was needed prior to wide-scale rollout:
I think it’s fantastic if it does work, because really if four hundred pounds during pregnancy does work fantastic to the cost of the NHS because you can have so much less extra scans during the pregnancy, complications during the pregnancy, less babies born premature, less still birth, less cot death, it’s going to cost the NHS thousands and thousands and thousands of pounds, millions really (FG15, I, smoking cessation advisors).
Cost savings due to reduced workloads for health service staff and how a ‘healthier workforce’, particularly in relation to smoking cessation, would result in employers having lower absenteeism rates were raised. Some professionals also stated how organisation incentives to meet targets could provide invaluable opportunities for needs-led service development; ‘to buy something for our service’.
From a more negative perspective, many participants believed that incentive programmes could create a ‘bad image’ and ‘negative publicity’ as a ‘waste’ of tax payers’ money and health service resources. In the current adverse economic climate, participants considered incentives to be the ‘wrong use of money’ when the health service was faced with ‘cut backs affecting visitation times’ and hospitals ‘trying to clear debts’ as well as the opportunity costs for resourcing other services:
If they [general public] thought they were actually getting money, just because there is so many cut backs, they may be seeing it that they’ve got an elderly relative with Alzheimer’s or something like that or nursing home fees (66, I, health improvement senior officer).
Others expressed how the increased expectations of payments for ‘other’ health behaviours, such as ‘obesity’, ‘healthy eating’, ‘alcoholics’ or ‘drug addictions’ could be a consequence:
Because if you start, obviously with the incentives with the breastfeeding, and they are smoking, well where do you draw the line (T11, I, mother).
The Cattaneo et al vignette [43] stimulated views from professionals that provider incentives were important to focus efforts on these behaviours and make sure ‘staff are more highly skilled’. CPIT providers reported how the initial ‘chaos’ and ‘resistance’ experienced when the incentive scheme was introduced ‘settled down overtime’. However, many professionals raised concerns about the associated costs of appointing committed, dedicated and skilled staff to deliver incentive programmes, and the potential implications of training, paperwork, administration, organisation, delivery and ‘policing’ of incentive delivery on ‘overstretched staff’ who were faced with ‘competing priorities’:
More staff input is a costly affair and staff are already under stress due to lack of time and resources so extra people would have to be recruited to implement these incentives and not just be added on to job descriptions (S164, midwife).
Discussion
To our knowledge, this is the first study to investigate the experienced and anticipated unintended consequences of incentive provision to either women or service providers for smoking cessation in pregnancy and breastfeeding. Findings highlight controversial and oppositional views towards financial incentives with ethical, political, cultural, social and psychological implications. We report how incentives can address or create inequalities; enhance or diminish intrinsic motivation and wellbeing, and how they may have a positive or negative impact on relationships within their personal networks and/or health providers. While incentives may raise awareness and direct service delivery to areas of need, this may be detrimental to other areas of health care. This detailed exposition thereby provides new insights into positive and negative consequences, as well as why, how and for whom these consequences might occur.
The strengths of this study include the mixed methods, multi-disciplinary, three site approach, which enabled us to engage with consumers and providers with and without direct experience of incentive interventions and the target behaviours. Data collection was undertaken by five researchers over a prolonged period of time. This resulted in variations in terms of how the interview and focus group questions were framed. Access to more disadvantaged settings and sampling techniques enabled us to obtain a broad range of views from participants with diverse socio-economic and behaviour characteristics, including participants who seldom engage in health services research. The limitations relate to where only restricted views were collected and the characteristics of the sample, despite every effort to recruit “harder to reach” participants. The White ethnic study population (88.6% of women/significant others and 89.3% of survey respondents) is slightly higher than census data for England (80.8%) [46] and Scotland (92.9%) [47]. In addition a much higher percentage of the women/significant others were married (77.3%) compared with rates for England in 2011 of (46.6%) [48]. A large percentage of the survey participants had never smoked (74.5%), which would be expected for health professionals. Furthermore only a small number of the women reported that they had not tried to breastfeed or planned to formula feed, perhaps reflecting that 80% of UK women initiate breastfeeding [49] or desirable response bias. In addition, limited views from wider family members were collected. Researchers selected the study vignettes from the evidence syntheses and drafted the shortlist of incentive strategies to frame the incentives. However, with service-user co-applicant input vignettes were also a strength, as the concrete scenarios enabled participants to highlight issues from a more individualised, reflective perspective than would have been achieved using more abstract question–answer techniques. A further limitation may relate to how the interviews/focus groups asked individuals to reflect on ‘consequences’ of incentives; whereas the survey specifically asked individuals to report on positive and negative consequences. In the light of our findings, more focused topic guides may well elicit additional insights, and this should be addressed within future studies. The limitations of the health professional survey are discussed fully elsewhere [14]. In particular, the free text question followed specific Likert Scale questions about agreement with our shortlist of seven most promising financial incentive strategies, which included shopping vouchers, a free breast pump and provider payments for meeting targets. Therefore it is possible that the free text responses could have been interpreted in the context of responses to the earlier questions. As non-response bias is also a concern, the free text data was only used for triangulation purposes.
Our findings on incentives to service providers support those previously reported for a wider range of health outcomes, in terms of how incentives can narrow health equalities through facilitating access to healthcare [4], [26]. While universal provision of incentives to consumers could guard against a postcode lottery of care, targeted support was important to engage those most at need and address embedded social norms. However, we identified the potential for differential uptake across social classes and the potential for health inequalities to increase, as noted for other lifestyle behaviour change interventions [50]. Of concern a MORI survey of the public acceptability of our shortlist of incentive strategies conducted as part of the BIBS study revealed important differences in attitudes between the more and less educated, and between women and men [14], [51]. Women compared to men were more likely to disagree with shopping voucher incentives for smoking cessation or breastfeeding, and those with lower levels of education, a reliable proxy for disadvantage [52], disagreed more with smoking cessation incentives and a breast pump [14], [51]. The assumption of governments that incentives will forge partnerships to deliver better health outcomes, reduce health inequalities and build social capital requires further testing.
At an individual level, incentives were the tipping point for some women, the opportunity that facilitated change [53]. However, as reported by others, concerns were raised that incentives could “crowd out” intrinsic motivation [4], [6], [7]. As incentives were identified to have the potential to enhance wellbeing, and wellbeing is identified as an important driver in behaviour-related infant feeding decisions [54], preferences for incentives that increase positive affect and add value for women as connectors for continued support [29] rather than diminish intrinsic motivation should be considered. How the consequences of incentives for behaviour change might impact either positively or negatively beyond the individual to social networks, communities and engagement within services is uncertain.
Media debates on the use of financial incentives have tended to be negative [55], [56]. However, a UK population based study by Promberger and colleagues to assess the acceptability of incentive-based treatments for smoking cessation and weight loss reported that the acceptability of financial incentives is not necessarily negative but rather contingent on the target behaviour, the type of incentive and their effectiveness [57]. Whilst positive unintended consequences have been reported in the literature [4], [18], [28], [29] our study also illustrates positive consequences that have received little attention to date. Autonomy is of particular note where incentives are perceived as bribes which can undermine free will and reflect a ‘nanny state’ resulting in diminished individual responsibility for health choices [58]. In contrast, consumers and professionals in our study reported that unrestricted vouchers can promote individual autonomy for the most disadvantaged through providing a rare opportunity for choice and self-reward. Similarly discourses of incentives as ‘unfair’ for rewarding ‘bad behaviour’, of discouraging individual responsibility or for targeting only disadvantaged communities contrast with narratives of feeling valued, more confident and improved self-esteem. Our interpretation is that media debates on the advantages and disadvantages of financial incentives can be easily biased towards intellectual philosophical, political and ethical arguments about the role of the state, without considering the perspectives of more disadvantaged families who are struggling to do the best for their children. This is particularly important as children have no choice and public acceptability is greatest for incentives to protect their wellbeing [2].
Related issues of ‘gaming and cheating’ were reported in terms of duplicitous activities amongst consumers and providers [17], [18], [25], [26], particularly when verifiable ‘proof’ of these behaviours was difficult to obtain. As incentives of higher values are believed to have a stronger correlation associated with unintended consequences [59], [60] consideration of the value and type (e.g. financial or behavioural) to mitigate against gaming behaviours appears crucial. The potential positive and negative impact of incentive schemes on provider–women relationships in our study are evident in the literature: increased access and improved rapport [29]; mistrust and alienation [4], [30] and limiting support for non-target behaviours [4], [5]. These findings highlight the need for sensitive, authentic, person-centred communication [61] as women dislike feeling judged or pressurised to behave in a way deemed appropriate by others [54], [62].
While incentives may have the potential to create shared aims, oppositional views emerged with regard to the impact of incentives on other health outcomes and service delivery. From one perspective, the investment in incentives would enable specialist and targeted services to be developed as reported by Cahill & Petera [18] and Gillam et al. [4]. However, the recent economic downturn led to views that investment in incentives was impractical, unethical and immoral, as well as creating expectations of ‘payments’ for other health-related behaviours [26]. Contentions largely centred on potential opportunity costs. However, in line with key government priorities of health promotion and prevention, a pro-active approach and associated funding to address these pervasive health behaviours were believed to have long-term benefits of reduced health costs, staff time and absenteeism rates.
A checklist has been developed to help decision makers assess when incentives might do more good than harm, to help prevent premature or inappropriate implementation [28]. This checklist highlights some of the key areas of contention in terms of ascertaining a) effectiveness of incentives prior to roll-out; b) appropriate targeting and eligibility criteria; c) valid and independent verification of behaviour outcomes; d) the implications of incentives on behaviours and motivation and e) that the benefits outweigh the potential for negative consequences in relation to ‘attention shift’ in terms of decreasing efforts in other health-care areas; ‘gaming’ behaviours and implications for the provider-consumer relationship. While this checklist highlights the need for systems and structures to be in place to prevent against negative consequences, our study emphasises how unintended positive consequences of incentive provision also require consideration.
Conclusion
The utility and acceptability of incentive provision is a controversial area, which can generate emotive and oppositional responses. Assumptions that incentives will help to address health inequalities, increase the reach of behaviour change interventions and facilitate a cultural shift towards desired behaviours require rigorous testing given the conflicting narratives around consequences. Prospective mixed methods approaches at the incentive intervention design and feasibility stages are needed, together with consideration of potential unintended consequences at all levels of service provision: for the participants, the population and those delivering the incentives and services. Process evaluations are also needed to capture the unintended consequences during a trial, with longer term follow up of key areas of concern after a trial has ended.
When planning an incentive intervention, care should be taken to assess discipline bias (e.g. philosophical, political, ethical, health) in narratives around personal autonomy, the role of the state and responsibility. These can be contrasted to personal narratives from less often heard voices representing the target population that incentives aim to help. Anticipation of positive and negative unintended consequences should be integral to the planning, design and implementation of interventions that include incentives, helping to ensure that any benefits are maximised.
Supporting Information
Appendix S1.
Web Survey for Early Years Professionals.
https://doi.org/10.1371/journal.pone.0111322.s001
(PDF)
Acknowledgments
We would like to thank all members of the wider research team, especially our co-applicant collaborators: Mastrick Café Crèche, Aberdeen and Wendy Ratcliffe who facilitated access; and Helen Cook at St Cuthbert’s and Palatine Children’s Centre, Blackpool. The data reported in this paper was collected as part of the BIBs study. The BIBs study authors are: Professor Pat Hoddinott, Professor Linda Bauld and Dr Stephan Dombrowski from the University of Stirling; Dr Heather Morgan, Graeme MacLennan, Kieran Rothnie, Dr Fiona Stewart, Dr Shelley Farrar, Dr Deokhee Yi, Dr Jenni Hislop, Dr Anne Ludbrook and Dr Marion Campbell from the University of Aberdeen; Dr Gill Thomson, Professor Fiona Dykes, Dr Victoria Hall Moran and Dr Nicola Crossland from the University of Central Lancashire; Dr Falko Sniehotta from the University of Newcastle and Professor David Tappin from the University of Glasgow. The lead author for this group is Professor Pat Hoddinott, University of Stirling, p.m.hoddinott@stir.ac.uk. We would also like to acknowledge that while Jennifer McKell and Susan MacAskill collected the qualitative data for the CPIT trial, they were unavailable to contribute to the writing of this paper.
We would also like to thank the members of the public, the women, families and staff from health services, local government, voluntary sector and other organisations, who generously provided their time by participating in the BIBS study. Additional thanks to colleagues at the University of Aberdeen, Fiona Stewart and Cynthia Fraser, for providing guidance with literature searching and reference management, and Lara Kemp for providing secretarial support.
Data Policy
The University of Aberdeen (study sponsors) and the HTA (study funders) actively encourage collaboration. Data excerpts are presented in the manuscript. The full dataset cannot be publically available for ethical reasons: public availability would compromise participant privacy. Additional supplementary de-identified quote material to support the findings and conclusions of the paper are available on request. Interested researchers can make an application to the PIs for the study: Prof Pat Hoddinott, (University of Stirling, p.m.hoddinott@stir.ac.uk) and Dr Heather Morgan (University of Aberdeen, h.morgan@abdn.ac.uk) to discuss data sharing and their data requirements. External users will be bound by a University of Aberdeen data sharing agreement, which will be in place prior to release, and will be required to work with the research team to comply with the ethics committee approval for this study. A University of Aberdeen data sharing platform will be used to enable researchers who are given permission to access the data in a secure format.
Department of Health disclaimer
This report presents independent research commissioned by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, MRC, CCF, NETSCC, the HTA programme or the Department of Health.
Author Contributions
Conceived and designed the experiments: GT HM NC LB FD PH. Performed the experiments: GT HM NC PH. Analyzed the data: GT HM NC LB FD PH. Contributed to the writing of the manuscript: GT HM NC LB FD PH.
References
- 1. Marteau TM, Ashcroft RE, Oliver A (2009) Using financial incentives to achieve healthy behaviour. BMJ 338: b1415.
- 2. Diepeveen SL, Ling T, Suhrcke M, Roland M, Marteau T (2013) Public acceptability of government intervention to change health-related behaviours: A systematic review and narrative synthesis. BMC Public Health 13: 756
- 3. Giles ER, S, McColl E, Sniehotta FF, Adams J (2014) The Effectiveness of Financial Incentives for Health Behaviour Change: Systematic Review and Meta-Analysis. PLoS ONE 9: e90347.
- 4. Gillam SJ, Siriwardena AN, Steel N (2012) Pay-for-performance in the United Kingdom: impact of the quality and outcomes framework: a systematic review. Ann Fam Med 10: 461–468.
- 5. Flodgren G, Eccles MP, Shepperd S, Scott A, Parmelli E, et al. (2011) An overview of reviews evaluating the effectiveness of financial incentives in changing healthcare professional behaviours and patient outcomes. Cochrane Database Syst Rev Issue 7: CD009255.
- 6.
Frey B (1997) Not just for the money: An economic theory of personal motivation. Brookfield, VT: Edward Elgar Publishing.
- 7. Deci EL, Koestner R, Ryan RM (1999) A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation. Psychol Bull 125: 627–668.
- 8. Johnston M, Sniehotta F (2010) Financial incentives to change patient behaviour. J Health Serv Res Policy 15: 131–132.
- 9.
Fiszbein AN, Schady N, Schady N, Ferreira FG, Grosh M, et al.. (2009) Conditional Cash Transfers: Reducing Present and Future Poverty. A World Bank Policy Research Report. 47603. Available: https://openknowledge.worldbank.org/bitstream/handle/10986/2597/476030PUB0Cond101Official0Use0Only1.pdf?sequence=1. Accessed 2013 Sep 5.
- 10. Merton RK (1936) The Unanticipated Consequences of Purposive Social Action. Am Sociol Rev 1: 894–904.
- 11.
Free Yourself - Smoking, It’s Not Worth It! (Tayside Quit4U Scheme). Available: http://www.canstopsmoking.com/local-help/free-yourself-smoking,-it’s-not-worth-it!-(tayside-quit4u-scheme). Accessed 2013 Oct 13.
- 12.
Encouraging Teenage Mothers to Breastfeed. Available: http://www1.uwe.ac.uk/bl/research/bsmc/researchprojects/healthandexercise/encouragingbreastfeeding.aspx. Accessed 2013 Sep 5.
- 13.
Breastfeeding mothers offered £200 in shop vouchers. Available: http://www.bbc.co.uk/news/health-24900650. Accessed 2013 Nov 12.
- 14.
Morgan H, Hoddinott P, Thomson G, Crossland N, Farrar S, et al. (in press) Benefits of incentives for breastfeeding and smoking cessation in pregnancy (BIBS): a mixed methods study to inform trial design. Health Technology Assessment http://www.nets.nihr.ac.uk/projects/hta/103102.
- 15. Hoddinott P, Hislop J, Morgan H, Stewart F, Farrar S, et al. (2013) Incentive interventions for smoking cessation in pregnancy: a mixed methods evidence synthesis. Lancet 380 (Suppl (3)) S48.
- 16. Cahill K, Perera R (2008) Competitions and incentives for smoking cessation. Cochrane Database Syst Rev Issue 3: CD004307.
- 17. Cahill K, Perera R (2011) Competitions and incentives for smoking cessation. Cochrane Database Syst Rev Issue 4: CD004307.
- 18.
Cahill K, Perera R (2008) Quit and Win contests for smoking cessation. Cochrane Database Syst Rev Issue 4: Art. No: CD004986.
- 19.
National Institute of Health and Clinical Excellence (2010) The use of incentives to improve health: Citizens Council Meeting. London, National Institute for Health and Clinical Excellence.
- 20. Cinciripini PM, Blalock JA, Minnix JA, Robinson JD, Brown VL, et al. (2010) Effects of an Intensive Depression-Focused Intervention for Smoking Cessation in Pregnancy. J Consult Clin Psychol 78: 44–54.
- 21. Van Herck P, De Smedt D, Annemans L (2010) Systematic review: effects, design choices, and context of pay-for-performance in health care. BMC Health Serv Res 10: 247.
- 22.
Kavanagh J, Trouton A, Oakley A, Powell C (2006) A systematic review of the evidence for incentive schemes to encourage positive health and other social behaviours in young people. Available: http://eppi.ioe.ac.uk/cms/LinkClick.aspx?fileticket=Soskzt1sFZs%3D&tabid=671&mid=1562. Available 2013 Feb 12.
- 23. Radley A, Ballard P, Eadie D, MacAskill S, Donnelly L, et al. (2013) Give It Up for Baby: Outcomes and factors influencing uptake of a pilot smoking cessation incentive scheme for pregnant women. BMC Public Health 13: 343
- 24.
Kavanagh J, Stansfield C, Thomas J (2009) Incentives to improve smoking, physical activity, dietary and weight management behaviours: a scoping review of the research evidence. Available: https://eppi.ioe.ac.uk/cms/Default.aspx?tabid=2456. Accessed 2013 Feb 12.
- 25. Kane RL, Johnson PE, Town RJ, Butler M (2004) Economic incentives for preventive care. Evidence Report: Technology Assessment (Summary) 101: 1–7.
- 26. Scott A, Sivey P, Ait OD, Willenberg L, Naccarella L, et al. (2011) The effect of financial incentives on the quality of health care provided by primary care physicians. Cochrane Database Syst Rev Issue 9: CD008451.
- 27. Paul-Ebhohimhen V, Avenell A (2008) Systematic review of the use of financial incentives in treatments for obesity and overweight. Obes Rev 9: 355–367.
- 28. Glasziou PP, Buchan H, Del M, Doust J, Harris M, et al. (2012) When financial incentives do more good than harm: a checklist. BMJ 345: e5047.
- 29. Thomson G, Dykes F, Hurley MA, Hoddinott P (2012) Incentives as connectors: insights into a breastfeeding incentive intervention in a disadvantaged area of North-West England. BMC Pregnancy Childbirth 12: 22
- 30. Coleman T (2010) Do financial incentives for delivering health promotion counselling work? Analysis of smoking cessation activities stimulated by the quality and outcomes framework. BMC Public Health 10: 167
- 31.
Edinburgh: Information Statistics Division. Breastfeeding statistics financial year 2011/12. NHS National Services Scotland. Available: http://www.isdscotland.org/Health-Topics/Child-Health/Infant-Feeding/. Accessed 2013 Oct 12.
- 32.
Edinburgh: Information Statistics Division. Data tables: births in Scottish hospitals - Year ending 31st March 2012. NHS National Services Scotland. Available: http://www.isdscotland.org/Health-Topics/Maternity-and-Births/Publications/data-tables.asp. Accessed 2013 Oct 12.
- 33.
Breastfeeding by NHS Board of Review: Financial Years 2001/02–2011/12. Edinburgh: Information Statistics Division, NHS National Services Scotland. Available: http://www.isdscotland.org/Health-Topics/Child-Health/Publications/data-tables.asp. Accessed 2013 Oct 12.
- 34. Tappin DM, Bauld L, Tannahill C, de Caestecker L, Radley A, et al. (2012) The cessation in pregnancy incentives trial (CPIT): study protocol for a randomized controlled trial. Trials 13: 113
- 35.
Child and Maternal Health Intelligence Network. Breastfeeding profiles. London, Public Health Available: http://atlas.chimat.org.uk/IAS/dataviews/report?reportId=351&viewId=355&geoReportId=3198&geoId=4&geoSubsetId=. Accessed 2013 Oct 2.
- 36.
Local tobacco control profiles for England. London, Public Health England. Available: http://www.tobaccoprofiles.info/tobacco-control#gid/1000110/par/E12000002/ati/102/page/0. Accessed 2013 Oct 2.
- 37. Chamberlain LB, McMahon M, Philipp BL, Merewood A (2006) Breast pump access in the inner city: A hospital-based initiative to provide breast pumps for low-income women. J Hum Lact 22: 94–98.
- 38. Gulliver SB, Colby SM, Hayes K, Raffa SD (2004) Tobacco cessation treatment for pregnant smokers: incorporating partners and incentives. Med Health R I 87: 9–12.
- 39. Heil SH, Higgins ST, Bernstein IM, Solomon LJ, Rogers RE, et al. (2008) Effects of voucher-based incentives on abstinence from cigarette smoking and fetal growth among pregnant women. Addict 103: 1009–1018.
- 40. Pugh LC, Milligan RA (1998) Nursing intervention to increase the duration of breastfeeding. Appl Nurs Res 11: 200–194.
- 41. Volpe EM, Bear M (2000) Enhancing breastfeeding initiation in adolescent mothers through the Breastfeeding Educated and Supported Teen (BEST) Club. J Hum Lact 16: 196–200.
- 42. Walsh RA, Redman S, Brinsmead MW, Byrne JM, Melmeth A (1997) A smoking cessation program at a public antenatal clinic. Am J Public Health 87: 1201–1204.
- 43. Cattaneo A, Borgnolo G, Simon G (2001) Breastfeeding by objectives. Eur J Publ Health 11: 397–401.
- 44.
Ritchie J, Spencer L (1994) Qualitative data analysis for applied policy research. In: Bryman A, Burgess RG, editors. Analyzing qualitative data. London: Routledge. pp. 173–194.
- 45. Braun V, Clarke V (2006) Using thematic analysis in psychology. Qual Res. Psychol 3(2): 77–101.
- 46.
Office for National Statistics (2011) Available: http://www.ons.gov.uk/ons/dcp171778_292378.pdf. Accessed 2014 Jul 30.
- 47.
Scotland’s Census (2011) Available: http://www.scotlandscensus.gov.uk/. Accessed 2014 Jul 30.
- 48.
Office for National Statistics (2011) Available: http://www.ons.gov.uk/ons/dcp171778_290685.pdf. Accessed 2014 Jul 30.
- 49.
McAndrew F, Thompson J, Fellows L, Large A, Speed M, et al.. (2012) Infant feeding survey 2010: summary. Available: http://www.hscic.gov.uk/catalogue/PUB08694. Accessed 2013 Aug 20.
- 50. Jepson RG, Harris FM, Platt S, Tannahill C (2010) The effectiveness of interventions to change six health behaviours: a review of reviews. BMC Public Health 10: 538
- 51. Hoddinott P, Morgan H, MacLennan G, Sewel K, Thomson G, et al. (2014) Public acceptability of financial incentives for smoking cessation in pregnancy and breastfeeding. BMJ Open 4: e005524.
- 52.
Marmot M (2004) Status Syndrome. London: Bloomsbury.
- 53.
Thaler RH, Sunstein CR (2008) Nudge: improving decisions about health, wealth, and happiness. New Haven: Yale University Press.
- 54. Hoddinott P, Craig L, Britten J, McInnes R (2012) A serial qualitative study of infant feeding experiences: idealism meets realism. BMJ Open 2: e000504.
- 55.
Reward healthy people with tax rebates to save the NHS, urges thinktank. Available: http://www.theguardian.com/society/2014/may/01/nhs-reward-healthy-with-tax-rebates. Accessed 2014 Jun 20.
- 56.
Park JD, Mitra N, Asch DA (2012) Public opinion about financial incentives for smoking cessation. Preventive Med 55 (Suppl): S41–S45.
- 57. Promberger M, Dolan P, Marteau TM (2012) “Pay them if it works”: Discrete choice experiments on the acceptability of financial incentives to change health related behaviour. Soc Sci Med 75: 2509–2514.
- 58. Ashcroft RE (2011) Personal financial incentives in health promotion: where do they fit in an ethic of autonomy? Health Expect 14: 191–200.
- 59. Doran T, Fullwood C, Kontopantelis E, Reeves D (2008) Effect of financial incentives on inequalities in the delivery of primary clinical care in England: analysis of clinical activity indicators for the quality and outcomes framework. Lancet 372: 728–736.
- 60. Sutton M, Elder R. Guthrie B, Watt G (2010) Record rewards: the effects of targeted quality incentives on the recording of risk factors by primary care providers. Health Econ 19: 1–13.
- 61. Schmied V, Beake S, Sheehan A, McCourt C, Dykes F (2011) Women’s perceptions and experiences of breastfeeding support: a metasynthesis. Birth 38: 49–60.
- 62.
Graham H, Sowden A, Flemming K, Heirs M, Fox D (2012) Using qualitative research to inform interventions to reduce smoking in pregnancy in England: a systematic review of qualitative studies: Public Health Research Consortium. Available: http://phrc.lshtm.ac.uk/project_2005-2011_a810.html. Accessed 2013 Feb 12.