SOCIAL WORK CASE STUDIES: CONCENTRATION YEAR

68

Social Work Research: Qualitative Groups

A focus group was conducted to explore the application of a cross-system collaboration and its effect on service delivery outcomes among social service agencies in a large urban county on the West Coast. The focus group consisted of 10 social workers and was led by a facilitator from the local office of a major commu- nity support organization (the organization). Participants in the focus group had diverse experiences working with children, youth, adults, older adults, and families. They represented agencies that addressed child welfare, family services, and community mental health issues. The group included five males and five females from diverse ethnicities.

The focus group was conducted in a conference room at the organization’s headquarters. The organization was interested in exploring options for greater collaboration and less fragmentation of social services in the local area. Participants in the group were recruited from local agencies that were either already receiving or were applying for funding from the organization. The 2-hour focus group was recorded.

The facilitator explained the objective of the focus group and encouraged each participant to share personal experiences and perspectives regarding cross-system collaboration. Eight ques- tions were asked that explored local examples of cross-system collaboration and the strengths and barriers found in using the model. The facilitator tried to achieve maximum participation by reflecting the answers back to the participants and maintaining eye contact.

To analyze the data, the researchers carefully transcribed the entire recorded discussion and utilized a qualitative data analysis software package issued by StatPac, which offers a product called Verbatim Blaster. This software focuses on content coding and word counting to identify the most salient themes and patterns.

The focus group was seen by the sponsoring entity as successful because every participant eventually provided feed- back to the facilitator about cross-system collaboration. It was also

RESEARCH

69

seen as a success because the facilitator remained engaged and nonjudgmental and strived to have each participant share their experiences.

In terms of outcomes, the facilitator said that the feedback obtained was useful in exploring new ways of delivering services and encouraging greater cooperation. As a result of this process, the organization decided to add a component to all agency annual plans and reports that asked them to describe what types of cross- agency collaboration were occurring and what additional efforts were planned.

O R I G I N A L R E S E A R C H

Process evaluation of a multiple risk factor perinatal programme for a

hard-to-reach minority group

Arlette E. Hesselink & Janneke Harting

Accepted for publication 5 February 2011

Correspondence to: A.E. Hesselink:

e-mail: [email protected]

Arlette E. Hesselink PhD

Post Doctoral Researcher

Department of Epidemiology and Health

Promotion, Public Health Service of

Amsterdam, The Netherlands

Janneke Harting PhD

Assistant Professor

Department of Social Medicine, Academic

Medical Centre/University of Amsterdam,

The Netherlands

H E S S E L I N K A . E . & H A R T I N G J . ( 2 0 1 1 )H E S S E L I N K A . E . & H A R T I N G J . ( 2 0 1 1 ) Process evaluation of a multiple risk

factor perinatal programme for a hard-to-reach minority group. Journal of

Advanced Nursing 67(9), 2026–2037. doi: 10.1111/j.1365-2648.2011.05644.x

Abstract Aim. This article is a report of an evaluation of a multiple risk factor perinatal

programme tailored to ethnic Turkish women in the Netherlands.

Background. The programme was directed at multiple risk factors and aimed at

improving maternal lifestyle, infant care practices and psychosocial health during

pregnancy and after delivery. The programme was carried out by ethnic Turkish

community health workers in collaboration with midwives and physiotherapists.

Methods. Our multiple case study included three Parent-Child Centres providing

integrated maternity and infant care. Participants (n = 119) were first and second

generation pregnant ethnic Turkish women with relatively unfavourable risk profiles.

Data were collected between 2005 and 2008 using mixed methods, including field

notes, observations and recordings of group classes, attendance logs, semi-structured

individual interviews, a focus group interview, and structured questionnaires.

Findings. Most participants (82%) were first generation ethnic Turkish; 47% had a

low educational level; 43% were pregnant with their first child; and 34% had a

minimal knowledge of the Dutch language. The community health workers’ Turkish

background was vital in overcoming cultural and language barriers and creating a

confidential atmosphere. Participants, midwives and health workers were positive

about the programme. Midwives also observed improvements of knowledge and self-

confidence amongst the participants. The integration of the community health

workers into midwifery practices was crucial for a successful programme imple-

mentation.

Conclusions. A culturally sensitive perinatal programme is able to gain access to a

hard-to-reach minority group at increased risk for poor perinatal health outcomes.

Such a programme may be well received and potentially effective.

Keywords: antenatal care, childbirth education, community health workers, hard-to-

reach community group, midwifery, perinatal, pregnancy, programme evaluation

� 2011 The Authors 2026 Journal of Advanced Nursing � 2011 Blackwell Publishing Ltd

J A N JOURNAL OF ADVANCED NURSING

Introduction

In many western countries, the largest differences in physical

health and mortality between migrants and host populations

are found during pregnancy and the first year of life

(Alderliesten et al. 2000, Gennaro 2005, Gagnon et al.

2009). Compared with the host population, ethnic minorities

have less favourable scores on various health indicators, such

as premature birth, lower birth weight, excessive infant

crying and sudden infant death syndrome (SIDS) (Steenbergen

et al. 1999, Schulpen et al. 2001, Alderliesten et al. 2008).

This increased health risk can be attributed in part to

modifiable risk factors with a relatively high prevalence

among first- and second- generation migrant mothers during

and after pregnancy. Among these are unhealthy lifestyles,

potentially harmful infant care methods (Schulpen et al.

2001, van Sleuwen et al. 2003, Goedhart et al. 2008),

depressive symptoms, pregnancy-related anxiety, perceptions

of racial discrimination and parenting stress (Huizink et al.

2003, Dole et al. 2004, van der Wal et al. 2007).

Background

In the Netherlands, risk factors for negative perinatal

pregnancy outcomes are addressed in part by midwifery

care. All pregnant women are advised to register with a

midwifery practice before the twelfth week of pregnancy and

to enrol in either a public or commercial perinatal pro-

gramme. Such programmes give information on healthy

pregnancy behaviours, childbirth and maternity care, and

prepare women for the approaching delivery and parenthood.

However, ethnic minority women rarely attend these perina-

tal programmes (Schulpen et al. 2001, Rowe & Garcia

2003). Apart from language and cultural barriers, this non-

attendance may be due to the fact that perinatal programmes

are not an integrated part of most ethnic cultures and are

entirely tailored to the native Dutch population. As these

programmes have the potential to effectively address risk

factors for negative perinatal pregnancy outcomes, and thus

to decrease the existing differences in health between host

and ethnic populations, there have been repeated recommen-

dations to develop perinatal programmes specifically tailored

to ethnic minorities (Schulpen et al. 2001, Alderliesten et al.

2008). Despite this high degree of interest, thus far only

minimal attention has been paid to developing and evaluating

perinatal programmes to reach minorities and change specific

risk factors (Ickovics et al. 2007, Joseph et al. 2009).

To fill this gap, the Public Health Service Amsterdam

(PHSA) has systematically developed a multiple risk factor

perinatal programme for first- and second-generation ethnic

Turkish women. This target group was chosen because of

increasing evidence of an accumulation of unhealthy lifestyle

factors and potentially harmful infant care behaviours

(Steenbergen et al. 1999, van Sleuwen et al. 2003, Goedhart

et al. 2009). The programme, which was called ‘Happy

Mothers, Happy Babies’ (HMHB), carried out by ethnic

Turkish community health workers (CHWs) in collaboration

with midwives and physiotherapists, and pilot-tested from

September 2005 to November 2007. This pilot implementa-

tion was accompanied by a process and effect evaluation.

The study

Aim

The aim of this study was to evaluate a multiple risk factor

perinatal programme tailored to ethnic Turkish women in the

Netherlands.

Design

The implementation process of the programme was evaluated

in a multiple case study. Data were collected between

September 2005 and October 2008 using a mixed-methods

approach (Creswell 2003).

Participants

We selected three of the seven out of fourteen Amsterdam

PCCs that had appointed an ethnic Turkish CHW to improve

perinatal care for ethnic Turkish women. This CHW had to

have a good command of the Dutch language so that she

could use the programme handbook and communicate with

the midwives, physiotherapists, programme coordinator and

researcher involved.

The midwives and CHWs were responsible for recruiting the

participants. The inclusion criteria were that the women

should be less than 24 weeks pregnant and first- or second-

generation Turkish. Eligible participants were invited for

individual consultations with the CHW and, after agreeing to

participate, grouped according to the expected date of delivery.

For logistical reasons, the initial plan was to start four

HMHB groups each year with a minimum of four and a

maximum of 15 participants per group. This was expected to

result in a sample size between 48 and 180 participants.

Conditions and participants were purposefully sampled

based on programme theme and CHW involved. Topic lists

were used for both observations and interviews. Data were

collected until theoretical saturation was reached (Polit &

Beck 2004).

JAN: ORIGINAL RESEARCH Process evaluation of a multiple risk factor perinatal programme

� 2011 The Authors Journal of Advanced Nursing � 2011 Blackwell Publishing Ltd 2027

Programme

To influence the distal outcomes on morbidity and mortality

level, the needs assessment revealed that the HMHB

programme should address the following determinants:

(1) smoking during pregnancy (Goedhart et al. 2008,

2009); (2) depressive symptoms (Karaçam & Ançel 2009);

(3) smothering, shaking and slapping of the baby (Reijne-

veld et al. 2004); (4) parent-child attachment (van der Wal

et al. 1998); and (5) healthy infant care practices such as not

smoking near the baby and using a blanket instead of a

duvet (van Sleuwen et al. 2003). In addition, three topics

subsistent in any perinatal programme were included: (6)

breastfeeding, which is normally good among the ethnic

Turkish population (van der Wal et al. 2001); (7) childbirth,

because of observed language and cultural barriers in

informing ethnic Turkish women (Crebas 2001); and (8)

getting maternity care, because of reported misunderstand-

ings about and suboptimal use of such services (Alderliesten

et al. 2008).

To effectively address these eight factors, evidence-based

and theory-based methods were included as much as possible

during an Intervention Mapping procedure (Bartholomew

et al. 2001). If appropriate, elements of existing perinatal

programmes were adapted to the ethnic Turkish target group.

Further tailoring of the programme was informed by the

practical experiences of key figures. The resulting programme

included various strategies for influencing the different

determinants. For example, education, skill-building activi-

ties and group discussions, a film to improve parent-child

attachment, role-plays to prevent smoking in the baby’s

immediate environment, and a game to promote healthy

infant care practices. A handbook, which included all

relevant information, was drawn up to give guidance for

the programme’s implementation (Jurg et al. 2005). The

handbook was written in Dutch and designed to be user-

friendly so that all information for each meeting was

structured around the topics that would subsequently be

addressed.

Table 1 gives an overview of the different topics and the

timetable. In total, the programme included eight group

classes of 2 hours each (seven before and one after delivery),

two individual consultations of 2 hours each (before deliv-

ery), and two home visits of 1 hour each (after delivery). Both

the mothers-in-law (first group class) and the partners

(second individual consultation) were invited once during

the course of the programme. These family members were

expected to have an influence on the infant care practices of

the mother (Turan et al. 2001, Geçkil et al. 2007) and to

sometimes play an essential role in permitting the women to

attend the programme. All but the first and the last group

class included 45 minutes of physical exercises, and infor-

mation about anatomy and the delivery. After delivery, a

‘baby show’ was scheduled: a final group class in which the

women could show their newborn babies to the other

participants.

Table 1 ‘Happy Mothers, Happy Babies’ perinatal education programme

Objectives

timetable

Preventing

smoking

during

pregnancy

Preventing

depressive

symptoms

Coping

with

the infant’s

crying

Increasing

parent-child

attachment

Promoting

healthy

childcare

practices

Preventing

smoking

near the

baby

Promoting

breast-

feeding

Preparing

for

childbirth

Requesting

maternity

care

Conception

After 14 weeks I1 � � . . . . . . . After 24 weeks G1 � . . . � � . . . # G2 . � . . . � . � � # G2 . � � . . . . . . # G2 . � . . . � � . . # I2 � . . � . � . . . # G2 . . . . . . . � � # G2 . . . . � . . � � # G2 . � � . � � . . .

Delivery

After 2–3 weeks H . . � . � � � . . After 8–10 weeks G3 . . . � . . . . . After 5 months H . . � . � � � . .

#, every week; I1, individual contacts; I2, individual contacts with partners invited; G, group class; G1, group class with mothers-in-law invited;

G2, group class including physical exercise with physiotherapist; G3, group class with ‘baby show’ (group class during which, in addition to

providing information, the participants get the chance to show their newborn baby to the other participants); H, home visit.

A.E. Hesselink and J. Harting

� 2011 The Authors 2028 Journal of Advanced Nursing � 2011 Blackwell Publishing Ltd

The HMHB programme was offered through three Parent-

Child Centres (PCCs), which give integrated maternity and

infant care. The programme was carried out by ethnic

Turkish CHWs and physiotherapists. The CHWs had already

received training in doing individual consultations and giving

group classes on health- and pregnancy-related issues for

minorities. They received an additional 2-day training during

which they received instruction on how to accurately

implement the programme in accordance with the pro-

gramme handbook. During the programme’s implementa-

tion, they were personally supervised by the programme

coordinator from the MPHS a midwife or PCC coordinator.

To enhance information exchange, the CHWs were

instructed to give the midwives with notes of each individual

participant contact. The CHWs were contracted in the PCC

for respectively 13 (District I), 12 (District II) and 8 (District

III) hours, depending on the number of ethnic Turkish

inhabitants living in the PCC district. The time available to

carry out the HMHB programme varied per PCC: 6 hours in

Districts I and II, and 4 hours in District III, which meant that

all CHWs also continued to be involved in the regular

maternity and infant care given by the PCC.

Data collection

The qualitative data collection existed of field notes, obser-

vations and recordings of group classes, attendance logs, semi-

structured individual interviews and a focus group interview.

Data were collected by the researcher and, if appropriate, a

research assistant. Triangulation of data and methods was

used to improve the credibility of the qualitative data.

The researcher summarized every contact with profession-

als in the field – such as the CHWs, midwives and PCC

coordinators – and noted interesting contextual information.

Observations were made during eight group classes, one

individual consultation and one home visit, after which the

applicable topics and strategies were discussed extensively

with the CHWs. Three group classes were translated into

Dutch to evaluate how accurately the programme had been

implemented. During the break and at the end of the group

class, the researcher spoke with participants about their

experiences with the programme, what they learned during

the programme and whether they valued it. The CHWs

additionally completed an attendance log after every individ-

ual consultation or group class, which included (a) whether a

participant was present or the reason for being absent; (b)

whether all topics were dealt with in accordance with the

programme handbook; and (c) special information about the

participant, for instance, whether she smoked, lived with her

parents-in-law or was depressed.

Individual semi-structured interviews were conducted with

midwives (one of each participating PCC), two CHWs (with

whom the researcher had the least contact), one PCC

coordinator (at whose PCC the implementation appeared to

be the most difficult), and the programme coordinator. Issues

of interest were the recruitment of participants, the imple-

mentation of the programme, the competencies required and

communication.

A focus group interview was held with the three CHWs

and the programme coordinator. During this interview, each

topic of the HMHB programme was discussed in terms of its

relevance, how it was elaborated and presented in the

programme handbook, and how it was presented to and

received by the participants.

The quantitative data were collected using structured

questionnaires. All Turkish pregnant women were inter-

viewed twice: the first time between 3 and 5 months after

conception (by the CHW), the second time 8 months after

conception (by a female Turkish interviewer). Information

was collected about the participants’ characteristics, such as

age, gender, ethnicity and knowledge of the Dutch

language. The second interview included five open ended

questions on how the participants had experienced the

programme, on what they had learned and on what was

lacking in the programme. No reliability checks were

performed for the quantitative demographic data; the open

ended questions served as method triangulation for the

qualitative data.

Ethical considerations

Committee approval was obtained from the Department of

Social Development, the Youth Health Care Institution and

the Public Health Service. Written informed consent from

all participants was obtained before the baseline measure-

ment.

Data analyses

All interviews were recorded, and transcribed verbatim. The

interviews were entered in MaxQDA and organized by topic.

Data from the focus group interview, observations and self-

administered notes were summarized manually. All qualita-

tive data were then subject to a content analysis (Polit & Beck

2004). To ensure inter-rater reliability, half of the data were

double coded by the researcher and the research assistant and

differences in coding were discussed until consensus was

reached. Descriptive statistics (frequencies and means) for the

quantitative information were obtained using the statistical

package SPSS 17 (SPSS Inc., Chicago, IL, USA).

JAN: ORIGINAL RESEARCH Process evaluation of a multiple risk factor perinatal programme

� 2011 The Authors Journal of Advanced Nursing � 2011 Blackwell Publishing Ltd 2029

Results

Participants

The HMHB programme ran 16 times in 2 years. Due to the

high workload in District III and the low number of

participants in District II, the CHWs were allowed to reduce

the number of HMHB programmes to three each year.

A total of 119 Turkish women started HMHB. The

questionnaires revealed that the mean age of the participants

was 27 years, that 83% were first-generation Turkish, and

that 40% were pregnant with their first child (Table 2). The

data recorded in the logbook showed that many women

attending the HMHB programme also had problems with

their partners, their parents-in-law and/or their financial

situation.

The number of women invited to participate was not

recorded systematically. According to the CHW, women who

refused to participate did not have time because they were

working, studying or had children at home; did not think

they needed a prenatal education programme; were not

allowed to attend by their husbands or parents-in-law; or felt

too Dutch to follow a course in Turkish.

Some Turkish women were more Dutch than Turkish and spoke the

Dutch language well, and were therefore not interested in HMHB.

(Midwife I)

Of the 119 women who had a first individual HMHB

consultation with the CHW, 105 (88%) actually took part

in one or more of the HMHB meetings (Table 3). Reasons

to drop out were: unclear (n = 7); moved elsewhere (n = 2);

busy with work or school (n = 4); or having other children

to care for (n = 1). The number of participants was the

lowest in District II (n = 24), the highest in District III

(n = 45) and somewhere in between in District I (n = 36).

This was not congruent with the number of ethnic Turkish

inhabitants and deliveries per month. One possible reason

for this was that the CHW was not integrated into

midwifery care.

Implementation sites

The HMHB programme was the easiest to implement in the

only PCC where both maternity and infant care were actually

available at the same location (District III). The integrated

nature of this PCC clearly improved the collaboration

between the midwives and the CHWs. This continuity was

lacking to a great degree in the other two districts.

I recognize that especially in District II, where the midwifery practice

is geographically further away from the main PCC, the contacts

between the different disciplines are more problematic than in the

other districts. (HMHB coordinator)

The integrated PCC in District III also allowed the CHW

to combine her work for the HMHB programme with her

regular work for both the midwives and the maternity care

workers in the PCC. This made her a familiar presence,

and as such, easy to approach for both clients and

professionals.

Collaboration and integration

Integrating the CHWs into the midwifery practices appeared

to be critical to implementing the HMHB programme. In two

districts, the CHWs were attached to the midwifery practice

and supervised by the midwives themselves (Districts I and

III). This made the midwives feel responsible for the CHWs,

enthusiastic about the CHWs’ role and performance and keen

about the HMHB programme itself. This resulted in a close

collaboration (i.e. regular face-to-face contact) in which the

Table 2 Sociodemographic characteristics of the HMHB partici- pants (n = 105)

HMHB

group, n (%)

or mean ± SDSD

Age (years) 27 ± 4Æ8 First generation* 87 (83)

First generation partner* 95 (81)

Length of residence in the Netherlands� 9Æ5 ± 7Æ5 Woman’s educational level�

Low 49 (47)

Moderate 24 (23)

High 32 (31)

Financial situation

Using savings or accumulating debts 49 (47)

Exactly enough 24 (23)

Good financial situation 32 (30)

Nulliparous (first child) 42 (40)

Smoking (yes) 20 (17)

Living with their parents (or parents-in-law) 19 (18)

Minimal knowledge of the Dutch language�,§ 36 (34)

Ethnic self-identification: feeling exclusively Turkish 85 (82)

*First generation: participant was born in Turkey; second generation:

one of her parents was born in Turkey. �First generation only. �Low: none or primary education; moderate: lower vocational

training and lower general secondary school; or high: intermediate

and higher vocational training, advanced secondary education and

university. §Combination of three questions about speaking, reading and

understanding the Dutch language.

A.E. Hesselink and J. Harting

� 2011 The Authors 2030 Journal of Advanced Nursing � 2011 Blackwell Publishing Ltd

CHWs were treated as an integrated part of midwifery care.

The CHWs were involved in additional tasks, such as

interpreting during midwife consultations, helping midwives

during home visits, and giving pregnant women information

on specific topics during extra consultations. In one district

(District III), the midwives even invited themselves to one of

the HMHB classes.

The most important reason we give information ourselves is to show

the participants that the CHW is part of our team and that we

support the information she is giving. (Midwife III)

In the other district (District II), the CHW was not

integrated into the midwifery practice. Most individual

consultations took place at the PCC, and some took place at

the midwifery practice on days when there were no midwife

consultations planned. For this reason, the communication

with the midwives remained limited mainly to the notes

recorded by the CHWs of their individual consultations

with the participants. Despite mediation efforts by the PCC

manager and the HMHB coordinator, the communication

and collaboration in this midwifery practice remained

problematic.

Recruiting participants and keeping them involved

The recruitment and continued involvement of participants in

the programme appeared to be the most difficult parts of the

intervention.

Recruitment is usually the biggest problem. When you want to invite

people for group classes, you tell people what it’s about, you explain

it again, and then someone else explains it. It’s something that needs

your constant attention and energy. (HMHB coordinator)

However, recruiting second-generation ethnic Turkish

women appeared easier than recruiting women from the first

generation.

Women from the second generation get in touch on their own and are

more independent, while women from the first generation are

reluctant. (CHW III).

At the start of HMHB, midwives asked the women to attend

the programme, and if they agreed, the CHWs invited them

for the first individual consultation. Since the number of

women approached was initially rather low, the CHWs in

Districts I and III started to collect patient records with

Turkish names and to invite these women by phone. Most

interviewees regarded the CHWs as being able to inform the

potential participants more effectively.

She (the CHW) can tell the women about the programme in their

own language, as well as why it is offered to them and what they can

expect. (Midwife III)

As in District II, the CHW was not an integrated part of

midwifery care, and she was not allowed to look for Turkish

names in the midwife’s patient records or to contact

women independently. Therefore, the recruitment of women

Table 3 Number of participants per district

Total District I District II District III

Total number of inhabitants* 41,335 43,913 33,847

% inhabitants with a Turkish background* 17 8 9

Absolute number of inhabitants with a Turkish background* 6806 3686 2914

Deliveries per month 38 25 27

Number of HMHB programmes given 16 6 5 5

Initial number of participants� 119 45 27 47

Participants who attended the programme� 105 (88%) 35 (78%) 24 (89%) 45 (96%)

Present for�

0–3 group classes 15 (14%) 8 (23%) 3 (13%) 3 (7%)

4 or 5 group classes 36 (34%) 13 (37%) 9 (38%) 14 (31%)

6 or 7 group classes 54 (