Development of a Scale for the Coordination Function of Family Physicians Regarding Antenatal Care

Research Article

Austin J Public Health Epidemiol. 2015;2(1): 1015.

Development of a Scale for the Coordination Function of Family Physicians Regarding Antenatal Care

Usturali Mut AN¹*, Öcek ZA¹, Çiçeklioğlu M¹ and Yücel U²

1Department of Public Health, Ege University Faculty of Medicine, Turkey

2Department of Midwifery, Ege University Izmir Atatürk School of Health, Turkey

*Corresponding author: Usturali Mut AN, Department of Public Health, Ege University Faculty of Medicine, 35100, Bornova-Izmir, Turkey

Received: January 14, 2015; Accepted: April 10, 2015; Published: April 13, 2015

Abstract

Aim: The aim of this study was to develop a scale to measure the level of coordination of antenatal care by primary care services in Turkey and to evaluate the reliability and validity of this scale.

Methods: The scale was developed in four steps. The first three steps were generating an item pool and conducting an expert panel and pilot study. In the last step, the scale was administered to 178 women living in three suburbs of Izmir, Turkey, who had given birth between November 2013 and February 2014. The split-half and Cronbach’s alpha tests were applied to assess internal consistency. Exploratory factor analysis was used to investigate construct validity.

Results: Six factors emerged from the factor analysis: accessibility and comprehensiveness of FHW services, accessibility and comprehensiveness of FP services, coordination of care by FPs, coordination of care by FHWs, FPs as first point of contact and recognising determinants of health. The factors explained 65.7% of the total variance. The Split-Half and Cronbach’s alpha reliability coefficients were 0.7 and 0.9 respectively.

Conclusion: The scale developed in this study is a specific tool aimed at evaluating the coordination of antenatal care and has successful psychometric features.

Keywords: Coordination; Antenatal care; Validity; Reliability; Family physician

Abbreviations

FP: Family Physician; FHW: Family Health Worker

Introduction

Fragmented delivery of health services and poor coordination are general problems acknowledged in many countries representing different types of healthcare systems [1,2]. The results of these problems are serious, and include medication errors, duplication of diagnostic procedures, progression of disease due to inadequate delivery of preventive services, conflicting medical regimes and recommendations to patients and higher costs [3,4]. The coordination function of primary care is considered to be the key to overcoming the disconnection between the different actors in the healthcare system. Coordination is basically concerned with how relevant actors in healthcare interact and communicate in regards to delivery of services [5] and how primary care providers integrate all aspects of care when patients must use other levels of health services [6,7].

The burden of incoordination is particularly important for antenatal care, given that pregnant women typically seek care from multiple providers and failure to integrate this process can cause duplication and omission of services [8,9]. Care coordination has been shown to improve birth outcomes, especially for underserved women [10]. Therefore, the development of a specific scale for pregnancy will give more guiding information about coordination during this period.

Turkey has a fragmented antenatal care system. One part of the system is primary care, where Family Physicians (FPs) work alongside a Family Health Worker (FHW) to provide individually-oriented services. FPs are general practitioners or family medicine specialists, whereas FHWs can be midwives, nurses or emergency medicine technicians. FPs have an obligation to conduct at least 4 check-ups at specified weeks during pregnancy for all the pregnant women on their lists. Other parts of the system are secondary and tertiary care services, which are provided by obstetricians. Secondary services include state hospitals, private hospitals whose services are partially reimbursed through the social security system and completely private hospitals, while tertiary care options are university hospitals. A pregnant woman can enter the system at whatever point she chooses without any limitations or referral system. There is no mechanism to inform FPs about the visits that are made in other settings [11,12]. Studies in Turkey have indicated that during pregnancy, women use many different sources of care at the same time [13-15] and FPs are not able to fulfill their coordination function [16,17].

The aim of this study was to develop a scale to measure the level of coordination of antenatal care by primary care services in Turkey and to evaluate the reliability and validity of this scale.

Methods

Development process of the scale

First step: An item pool was generated by taking items from the scales measuring four cardinal functions of primary care [18-21]. In the selection of items, weight was given to the coordination function of primary care, but items evaluating other functions necessary for coordination, such as first-contact, comprehensiveness and continuity, were also included. Items used in studies evaluating the coordination of prenatal care identified in a literature review were also added to the pool [22-25]. Subsequently, the research team chose 56 of the 88 items from this pool and restated them with particular reference to prenatal care.

Second step: In order to determine the face validity of the initial draft of the scale, an expert panel was organised comprising of three public health specialists, three FPs and two FHWs. After this panel, the number of questions was reduced to 36 and some questions were restated. It was decided that the questions should be answered on a 5 point Likert Scale (5=always, 4=usually, 3=sometimes, 2=rarely, 1=never).

Third step: The scale was pre-tested on a small sample (n=16), stratified by educational level. These women did not participate in the fourth step. After this pilot study, amendments were made in terms of wording and order of the items. Six items which were noticed not to have been understood were removed.

Fourth step: The psychometric properties of the final version of the scale (30 items) were tested in three poor suburbs (Mevlana, Naldöken, Altindağ) of Izmir, where the municipality of Bornova district is providing a social care program. The aim was to reach all women who lived in these suburbs and had given birth between November 2013 and February 2014 (n=248) during a dental health education programme organised by the municipality. 71.8% (n=178) of the women participated in the education programme and all of these mothers were included in the present study.

Statistical Analysis

The means and standard deviations of participants’ responses to each item were calculated. The Cronbach’s alpha coefficient test and split-half technique were applied to assess internal consistency. With split-half reliability, the items in the scale were divided into two groups and the relationship between respondents’ scores for the two halves was computed [26].

Exploratory factor analysis was used to investigate construct validity. Principal components analysis, the most widely used form of factor analysis, was performed to analyse all the variance of a variable, including its unique variance. Varimax rotation of orthogonal rotation methods was chosen in order to produce factors which were unrelated to or independent of one another [27].

The results of the Kaiser-Meyer-Olkin (KMO) and Bartlett’s test of sphericity were determined to measure sampling adequacy. Items with MSA (Measures of Sampling Adequacy) values of less than 0.5 and/or factor loadings of less than 0.4 were removed from the analysis [28]. In considering the Kaiser’s criterion, the number of initial eigenvalues higher than one was used for deciding on the number of factors [26,28]. The SPSS statistical program (SPSS Statistical Package® 21.0, IBM Corporation, 2012, Armonk, NY, USA) was used for the analysis.

Results

The average age of the women was 28.3±5.1 (minimum: 19, maximum: 42). The distribution of participants according to some descriptive variables is shown in Table 1. According to this, one in seven women was illiterate and one in four women had not completed primary education. The native language of one in three women was not Turkish and one in four women had no social security.