III. 2.2.7.3 Parámetros sanguíneos y urinarios
III.5. DISCUSIÓN
III.5.1. FERMENTACIÓN RUMINAL
This section reports the results of the evidence synthesis for incentives provided to organisations to
improve smoking cessation in pregnancy and breastfeeding outcomes. Three studies have been included in
this section of the review. Two provided incentives to organisations– one reimbursed an organisation for
the provision of smoking cessation counselling152and one provided a financial incentive for breastfeeding
promotion168– and one collected survey data, which included attitudes towards reimbursement of
health-care providers in relation to providing a smoking cessation intervention.153The studies that gave
incentives to care-providing organisations are summarised in Table 32.
TABLE 32 Summary of studies included in the review of incentives provided to organisations for smoking cessation
in pregnancy and breastfeeding
Study Country
Study design as
described Total n participants Incentive Primary outcome
Smoking cessation Latts
2002152
USA Pilot project to
implement a new smoking in pregnancy counselling system in a health maintenance organisation 18 physician practices representing 27 office sites and 80 physicians. Training was provided to 66 staff members
US$150.00 for each pregnant smoker counselled
Reimbursement received for providing counselling to four of 21 smokers identified
Hartmann 2007153
USA Survey of prenatal
care providers about smoking cessation interventions. One of 18 questions asked about the influence of reimbursement on providers’ willingness to provide a smoking cessation intervention 844 health professionals consisting of obstetricians (50%), midwives (18%), family medicine physicians (15%), nurse practitioners (13%) and physician assistants (4%)
No incentive provided
Over half of providers reported that reimbursement was at least‘somewhat’ influential in relation to their willingness to provide smoking cessation services, with no difference between provider types Breastfeeding Cattaneo 2001168 Italy Breastfeeding promotion programme
Six local health authorities, 18 public hospitals
Financial penalty for not reaching a self-set breastfeeding prevalence target
Smoking cessation in pregnancy
The smoking cessation review identified one study that had considered the effect of incentivising a
care-providing organisation.152The intervention involved training 66 staff in smoking cessation counselling
in obstetric practices delivering health maintenance organisation (HMO) care. The HMO was billed US$150.00 per woman, in increments of US$25.00 per 10 minutes, for providing counselling to women. Only 18 of the 33 obstetric practices agreed to participate. Before and following the intervention, reviews of medical charts of patients were undertaken to collect data on the identification of smoking status and the provision of advice to stop smoking. Smoking status was identified in 96% (175/182) of patients post intervention compared with 90% (198/220) of patients before the intervention. Smoking status identification was carried out at the first obstetric visit for 99% (196/198) at baseline and in 100% of cases (171/171) after the intervention. For the 12% (21/170) of women currently smoking in pregnancy at the time of the survey, documentation of advice to stop smoking actually worsened, from a baseline of eight out of 13 (62%) to four out of 17 (24%) after the intervention. The HMO received claims for reimbursement for only four out of 21 (19.1%) identified smokers who received counselling over the study period.
Patients were asked to recall components of the care provided to them in a follow-up survey, which had a response rate of 59% (505/849) at baseline and 33% (250/753) after the intervention. For the respondents who reported being current or former smokers, 96.8% (120/124) recalled being asked whether they were a smoker after the intervention compared with 90.2% (74/82) at baseline. However, only 64.5% (20/31) recalled being advised to quit by their obstetric provider after the intervention compared with 85.9% (61/71) before the intervention. After the intervention 3.2% (1/31) recalled
receiving smoking cessation counselling compared with 11.3% (8/71) before the intervention. At both time points no participant could recall being referred to a smoking cessation programme.
Staff acceptability for eligible staff still in employment was reported for 56.9% (29/51) who responded to a post-intervention survey. Of the 29 respondents, 23 indicated that they had used their training to provide smoking cessation counselling, three reported that they did not see any smokers to counsel, five did not see any smokers from the participating HMO, one reported having no interest in counselling smokers and one felt unqualified to counsel smokers despite training. The authors identified several barriers to participation and engagement in the intervention, including the presence of office managers who themselves smoked. They suggest that only the more enthusiastic physicians participated, yet despite this the intervention resulted in a decrease in the provision of counselling.
Prenatal care provider organisation attitudes were captured in a survey about best practice in providing a
smoking cessation intervention.153Within a mailed survey of 18 items, one question asked about‘the
influence of reimbursement on providers’ willingness to provide smoking cessation intervention’, with
respondents rating their willingness on a Likert scale from‘none’ to ‘very substantial’. Numerators and
denominators were infrequently reported in this study. More than half of the 844 respondents, who comprised obstetricians (50%), midwives (18%), family medicine physicians (15%), nurse practitioners (13%) and physician assistants (4%) from an area in the USA, reported that reimbursement was ‘somewhat’ influential on their willingness to provide smoking cessation counselling, with no difference between provider types. Of those who thought that reimbursement would be influential, 40% were less likely to provide best practice. The most commonly reported barriers to providing smoking cessation services were time and a perceived lack of patient interest. Appropriate reimbursement was recommended for further investigation.
Breastfeeding
by our search strategy and we did not identify any studies specifically investigating the incentive component of BFI among the 22 studies identified and hence only a brief overview is provided below.
Beake and colleagues184conducted a systematic review to assess whether a structured programme such as
the WHO/UNICEF BFI implemented in maternity acute care settings is more likely to be associated with higher rates of initiation and duration of exclusive breastfeeding than no structured programme. The ‘structured programme’ included a multifaceted approach to support breastfeeding that targeted change at organisational, service delivery and individual behaviour levels. Studies that considered only
community-based interventions were excluded. An extensive search of literature published in 1992–2010
was undertaken and methodological quality was assessed using checklists developed by the Joanna Briggs Institute. Twenty-six studies were included: one cluster RCT, two controlled trials, one cross-sectional study, two descriptive studies, 15 cohort studies and five systematic reviews.
Because of clinical and methodological heterogeneity of study designs it was not possible to combine studies or individual outcomes in meta-analyses. Most studies found a statistically significant improvement in breastfeeding initiation following introduction of a structured breastfeeding programme, although effect sizes varied. An impact on the duration of exclusive breastfeeding and duration of any breastfeeding to 6 months was also evident, although not all studies found statistically significant differences. Few studies controlled for any potential confounding factors, and the impact of bias has to be considered. It was concluded that structured programmes compared with standard care positively influence the initiation and duration of exclusive breastfeeding and any breastfeeding. In health-care settings with low breastfeeding initiation and duration rates, structured programmes may have a greater benefit. The generalisability of this evidence to the UK context is still debated as, to date, increased breastfeeding duration beyond the first
week after birth is uncertain.210This is perhaps not surprising given that the effects of the more recently
implemented BFI in the community, which aims to sustain breastfeeding after hospital discharge, have not been reported.
Cattaneo and colleagues168conducted an observational study in Italy of a breastfeeding promotion
programme that attributed financial incentives to local health authorities (LHAs) for complying with work plans and achieving targets set by the LHAs themselves. Financial penalties were levied against LHAs that did not comply with or did not achieve the targets. None of the hospitals in the LHAs had been designated ‘baby friendly’. The breastfeeding work plans and targets varied across the LHAs and included process or activity objectives, such as policy development, training of health professionals, education of mothers and improving co-ordination and integration of teams and activities, and outcomes measures, such as
increasing breastfeeding rates. The penalty for not achieving their own work plans and targets or not participating in the collection of data was a fixed percentage (0.5 per thousand of the amount paid by the
region to the LHA every year) (Adriano Cattaneo, Istituto per I’lnfanzia, Trieste, Italy, May 2013,
personal communication).
Ten hospitals across six LHAs set up a breastfeeding reporting system in 1998 and defined breastfeeding promotion activities for 1999. High variability in breastfeeding outcomes at hospital discharge was
recorded among LHAs. Overall, between 1998 and 1999, there was a statistically significant trend towards
an increased exclusive breastfeeding rate at discharge (from 37% to 49%; p< 0.001), although two LHAs
showed either no difference or a reduction in the exclusive breastfeeding rate. The overall rate of
predominant breastfeeding at discharge decreased (from 46% to 39%; p< 0.001) but full breastfeeding
rates did not change in four of the six LHAs. The rate of complementary breastfeeding at discharge also
declined from 15% in 1998 to 10% in 1999 (p< 0.001).
The variability in breastfeeding outcomes among LHAs was much smaller at 16–19 weeks than at hospital
discharge. Overall, there was a small but significant increase in exclusive breastfeeding rates between 1998
and 1999 (from 26% to 30%; p< 0.001) with an associated increase in the rate of full breastfeeding
(from 38% to 41%; p< 0.001) and a decrease in the rate of complementary breastfeeding (from 25% to
Published data are not available for subsequent years but as all six LHAs achieved their targets for every
year that the programme was active (1998–2004) no penalties were applied (Adriano Cattaneo, personal
communication). The data suggest an association of a financial incentive mechanism with improved breastfeeding rates. As the financial mechanism contributed to the establishment of a reporting system for breastfeeding and to the development of work plans and targets, the effect may result from the
introduction of a surveillance system with regular feedback or the modified practices carried out in each LHA to achieve the objectives, many of which are known to be effective in improving breastfeeding outcomes. Although this intervention shows promise, there was a sharp fall in breastfeeding rates at
hospital discharge compared with around 4 months (16–19 weeks), suggesting that changing hospital
practices is not sufficient to prolong breastfeeding, even when initiation is high.
Discussion
Incentives for smoking cessation in pregnancy
In relation to smoking cessation, there is good evidence that pregnant women provided with contingent voucher incentives with biochemical validation of outcomes are more likely to stop smoking by the end of pregnancy than those who receive non-contingent incentives (for participation in a smoking cessation programme and providing outcome data) alone. This is shown by the four studies meta-analysed for
end-of-pregnancy data in our results.105,109,141,144From the post-partum data available from three of the four
studies meta-analysed the difference in abstinence rates between the groups is carried into the early
months of the postpartum period (≤ 3 months) and is statistically significant. However, by 6 months
post partum, data available from two studies indicate that the difference is no longer statistically significant. The meta-analysis provides some information on the effect of varying the level and type of incentives. However, the evidence on incentives more generally is varied. At the end of pregnancy, cessation rates
ranged from 13.4% to 71.0% for incentivised groups compared with 6.0–60.0% for control groups
provided with smaller non-contingent incentives for participation. Intervention groups achieved abstinence
rates of 10.2–42.0% for the early post-partum period (from 0 to 2 months) whereas control groups
achieved abstinence rates of 0.8–38.0% for the same period. At 6 months post partum, intervention
groups and control groups achieved cessation rates of 36.0–37.0% and 0–33.0% respectively. The
findings of our meta-analysis are consistent with the findings of others.54,211A study by Sexton and Hebel212
was included in an earlier meta-analysis of incentives for smoking cessation in pregnancy54but was
excluded from our review. The abstract was not picked up through our search and the full-text paper reporting the trial outcomes does not mention incentives or rewards at all. It does highlight that intervention strategies were reviewed throughout the study, and new ideas and approaches were
incorporated. The paper was similarly excluded from a recent NICE review of the incentives evidence.41
Reviews by Higgins and colleagues211and Sigmon and Patrick213did not include any new eligible
published studies.
Aside from meta-analysed data described above, reported rates of abstinence included maximum values for incentivised groups that exceeded the maximum values for control groups, but there is overlap in the ranges and it is very difficult to draw any firm conclusions about the effectiveness of incentives from these data, particularly as the other components of the interventions being provided also varied widely.
It was also difficult to consider the effectiveness of interventions when studies did not have a comparator (non-incentivised) group. Cessation rates for these studies ranged from 13.6% to 50% at the end of pregnancy, from 21.3% to 37.0% during the early post partum period, from 7.9% to 77.0% at 3 months post partum and from 7.0% to 64.0% at 6 months post partum.
lack of consistency in the field as to how smoking cessation in pregnancy should be measured. Although
for general population smoking studies the Russell standard163is often used, providing some consistency,
a similar standard is not currently available for studies involving pregnant women.
Biochemical verification of smoking cessation was common, although the methods for determining abstinence and the cut-offs used varied and were problematic. It could be argued that standards for biochemically validating smoking cessation may not be adequate when incentive payments to individuals are made. CO and cotinine levels are both likely to be lower among pregnant smokers than among
non-pregnant smokers because of more rapid clearing of these metabolites of smoking.121Before the
results of the studies described above are accepted as true, there needs to be further study of pregnant
smokers’ ability to ‘game’ the system to receive incentive payments while still smoking. One way to look at
CO biochemical verification would be to make it clear that incentives are paid on self-report corroborated by CO level, but to take a cotinine level at the same time to assess gaming. A second method would be to use blood samples that are taken routinely at an appropriate time in pregnancy for other reasons (i.e. at
36 weeks’ gestation). These routine blood samples could be used to assess the extent of gaming at the
later pregnancy assessment for smoking cessation (corroborated by cotinine or CO level). It would be prudent to accept the findings with reservations until it is clear that gaming is minimal among pregnant smokers who are offered financial incentives for smoking cessation.
Some studies reported outcomes other than smoking cessation. Although some reported reductions in
cigarette consumption,104,138,140,149study design makes it difficult to determine whether or not the incentive
component of the interventions prompted the reductions.
Incentives for breastfeeding
With regard to breastfeeding, there is currently insufficient evidence to formulate conclusions on the effectiveness of incentives to improve breastfeeding outcomes. The lack of good-quality RCTs prohibited meta-analysis and the heterogeneity of the studies limits comparisons. The studies included in this review were different in a number of aspects, such as the type and complexity of the intervention, the comparison groups, study design, sample size, outcome definitions, follow-up times and the population studied. In total, 13 category B studies (see Table 21) compared an incentive of any type with no incentive or a much smaller incentive. Breastfeeding rates at initiation or at hospital discharge ranged from 38.1% to
88.5% in incentivised groups and from 14.6% to 72.2% in non-incentivised groups.103,107,175,176,179,181
The average duration of exclusive breastfeeding ranged from 0.7 to 12 weeks for incentivised groups and
from 2.78 to 12 weeks for non-incentivised groups.170,171,174,180The average duration of any breastfeeding
ranged from 4.0 to 136.3 weeks for incentivised groups and from 15.7 to 88.3 weeks for non-incentivised
groups.106,167,171,174Reported breastfeeding initiation and duration rates in intervention and control groups
therefore showed considerable overlap, limiting the ability to formulate conclusions about the effectiveness of incentives from these data.
The most commonly used incentive involved providing access to a breast pump. Seven studies evaluated the effect of provision of only a breast pump on breastfeeding outcomes and one evaluated the effect of the provision of a breast pump plus other gifts and vouchers. Of the studies that investigated breast
pumps only, four RCTs,167,170,172,180ranging in size from 34 to 1625 women, provided little consensus.
As statistically significant findings were identified only when women given breast pumps were compared
with those given infant formula,167,170it is not possible to ascertain whether it was the incentive of a breast
pump to breastfeed or the incentive of infant formula to formula feed that produced the observed effect. When breast pumps were provided with a number of other incentives such as gift items, vouchers and
raffle prizes in a small RCT of 55 women,176the proportion of women who exclusively breastfed was
significantly higher in the intervention group than in the control group (who received usual care and small
incentives). The remaining studies evaluating breast pumps included a conference abstract180and two
observational studies103,169and so the results should be treated with caution. In summary, studies
restrictions on breast pump availability, contamination of intervention and control groups, short follow-up
times and lack of adequate control groups. A Cochrane review214(currently withdrawn) investigated the
influence of the provision of commercial discharge packs that include infant formula samples and/or infant formula promotional materials on the duration and exclusivity of breastfeeding. Commercial discharge packs reduced the number of women exclusively breastfeeding at all times from 0 to 6 months but had no significant effect on non-exclusive breastfeeding. There are also ethical issues concerning the provision of infant formula incentives, which would contravene the WHO International Code of Marketing of
Breast Milk Substitutes.80Given the reported problems with intervention fidelity,174there is a need for an
adequately powered cluster RCT that compares breast pump provision with usual care to clarify the impact of breast pumps on breastfeeding rates.
Overall, it is difficult to draw any conclusions about the effectiveness of incentives from these data, particularly as the incentives differed and most interventions consisted of a number of BCT components. For example, most studies incorporated an education and/or a support element in their intervention, with the incentive provided either to encourage continuation in the programme or as a reward for continuing to breastfeed.106,107,171,173,175,176,178,179
It was also difficult to consider the effectiveness of interventions when studies did not have a adequate comparator (non-incentivised) group. A number of studies either did not include a non-incentivised control group169,172,182or used historical control subject, which may be subject to strong selection bias.103,107,112,175,179,181
Incentive intervention delivery for smoking cessation and breastfeeding