Health professionals’ experiences and views on obstetric ultrasound in Vietnam: A regional, cross-sectional study

Objectives Obstetric ultrasound is an important part of antenatal care in Vietnam, although there are great differences in access to antenatal care and ultrasound services across the country. The aim of this study was to explore Vietnamese health professionals’ experiences and views of obstetric ultrasound in relation to clinical management, resources and skills. Design A cross-sectional questionnaire study was performed as part of the CROss Country UltraSound study. Setting Health facilities (n=29) in urban, semiurban and rural areas of Hanoi region in Vietnam. Participants Participants were 289 obstetricians/gynaecologists and 535 midwives. Results A majority (88%) of participants agreed that ‘every woman should undergo ultrasound examination’ during pregnancy to determine gestational age. Participants reported an average of six ultrasound examinations as medically indicated during an uncomplicated pregnancy. Access to ultrasound at participants’ workplaces was reported as always available regardless of health facility level. Most participants performing ultrasound reported high-level skills for fetal heart rate examination (70%), whereas few (23%) reported being skilled in examination of the anatomy of the fetal heart. Insufficient ultrasound training leading to suboptimal pregnancy management was reported by 37% of all participants. ‘Better quality of ultrasound machines’, ‘more physicians trained in ultrasound’ and ‘more training for health professionals currently performing ultrasound’ were reported as ways to improve the utilisation of ultrasound. Conclusions Obstetric ultrasound is used as an integral part of antenatal care at all selected health facility levels in the region of Hanoi, and access was reported as high. However, reports of insufficient ultrasound training resulting in suboptimal pregnancy management indicate a need for additional training of ultrasound operators to improve utilisation of ultrasound. Since 2010, Vietnam has been classified as a lower middle-income country and has undergone substantial economic development in recent decades.1 The maternal mortality rate has decreased from 139/100 000 in 1990 to 54/100 000 live births in 2015.15 Most inpatient healthcare is provided by public hospitals, but for outpatient care private clinics account for a large number of patients.16 Vietnam is divided into 63 provinces, 698 districts and 11 121 communes.17 In each commune, a village health worker (VHW) provides health promotion, immunisation and nutrition services, and attends births in remote areas.18 At the community health centre level, a midwife or an assistant doctor is in charge of maternal health services, and provides ANC, assists normal delivery, and provides postnatal care, immunisation services and supervision of the VHW. At the district level, ANC, delivery care including caesarean sections and newborn care are provided at hospitals,18 while maternity homes deliver basic prenatal and delivery services.16 18 Provincial hospitals provide more specialised healthcare, and referrals from lower healthcare levels to provincial level are undertaken if complications occur during pregnancy or delivery.18 At the top of the healthcare system, there are several national hospitals providing specialised care and receiving referrals from lower levels.16 This cross-sectional study used a questionnaire to investigate a number of research questions related to obstetric ultrasound with obstetricians/gynaecologists and midwives providing pregnancy, delivery and postpartum care to women in the region of Hanoi, Vietnam. Owing to the lack of findings from similar studies, a sample size of 290 obstetricians/gynaecologists and a corresponding number of midwives (n=290) was calculated based on plausible estimations of prevalence of background characteristics and outcome variables. The calculation was based on the outcome requiring the largest sample size, ‘every woman should undergo ultrasound examination in pregnancy to determine gestational age’, and the background variable ‘work experience over and under 5 years’, to detect a difference in proportion of 0.10 with a power of 80% and a significance level of 5%. Purposive sampling was used to obtain a representative sample of health professionals caring for pregnant women at different levels of health facilities in urban, semiurban and rural areas in the region of Hanoi. One national hospital, 1 provincial hospital, 24 district hospitals and 3 maternity homes were included in the study, for a total of 29 health facilities. The study questionnaire was developed based on the results from the earlier qualitative studies performed in the CROCUS study.7 19–27 Sociodemographic characteristics, evaluation of self-reported skills in performing ultrasound, and questions about access to obstetric ultrasound and health professionals’ views on what may improve utilisation of ultrasound in Vietnam were included, among other items. The questionnaire included both questions and statements, and the items had either fixed or Likert-scale response options. This analysis investigates the research questions noted above, using 45 of the 105 questions and statements. The questionnaire was developed first in English and thereafter translated to Vietnamese by a native Vietnamese speaker independent of the research team. The Vietnamese version of the questionnaire was also back-translated to English by another external person. This check resulted in minor adjustments of some words, but demonstrated that the Vietnamese translation had retained the overall meaning of the English. The questionnaire was pilot-tested with 10 obstetricians, 6 midwives and 2 sonographers. No further revisions of the questionnaire were required as a result of piloting. The data collection was performed in April 2017 by four experienced data collectors supervised by two Vietnamese senior researchers in the research team. Before the start of the data collection, data collectors were trained by the research team, and all questions and statements in the questionnaire were discussed to ensure correct understanding. The two Vietnamese researchers initiated contact with the directors of all selected health facilities and all of them agreed to assist with recruitment of participants. For this study, we aimed to include health professionals caring for pregnant women and with different experiences in relation to use of obstetric ultrasound. Eligible participants were health professionals managing pregnant women at the maternity wards on the day of data collection at each study site. No eligible participant declined participation in the study. The primary sample included 890 participants. Six individuals working as radiology technicians were excluded from the primary sample as the they did not fulfil the inclusion criteria, and finally 60 sonographers were also excluded from the primary sample since they constituted a small part of the sample, and further did not contribute to clinical management after their obstetric ultrasound examination. The final sample (N=824) included the following health professionals: obstetricians/gynaecologists (n=289) and midwives (n=535). Participation was anonymous and all questionnaires were given a unique code. Safe storage of questionnaires was undertaken in accordance with national procedures and regulations. Data were entered into an SPSS file at Hanoi Medical University, by two experienced data clerks. To evaluate the quality of the data entry, every 10th questionnaire based on the number order of identification codes was selected for data re-entry. The data from all 107 variables in 89 questionnaires were re-entered in the SPSS file by the first author. The rate of error was 1.4%. The identified errors in the SPSS file were corrected. This research was done without patient involvement. Age was calculated as a continuous variable using birth year and year of data collection. For some analyses age was dichotomised as age 34 years or less and 35 years and above. Gender included female or male. Health profession included the following response options on the questionnaire: obstetrician/gynaecologist, general practitioner, resident physician, physician other (please specify), midwife, radiologist/sonographer and ‘other’ (please specify). Health profession was thereafter categorised into two groups: obstetricians/gynaecologists and midwives. Resident physicians undergoing postgraduate training (n=9) and general practitioners (n=12) were also included in the category obstetricians/gynaecologists because they worked at the same department and performed similar work tasks as the obstetricians/gynaecologists. One participant who was an anaesthesiologist by profession but was working with maternity care was categorised as an obstetrician/gynaecologist. One nurse working in maternity care was categorised as a midwife. Health facilities included the response options national hospital, provincial hospital, district hospital and maternity home. The variable health facilities was dichotomised into national hospital/provincial hospital and district hospital/maternity home in some analyses. Area of health facility was categorised as hospitals in urban (n=7), semiurban (n=5) and rural (n=17) areas of Hanoi. Type of healthcare was classified as public, private, and both public and private healthcare. No participant reported working only in private healthcare. Number of ultrasound examinations indicated in an uncomplicated pregnancy was categorised based on the three recommended number of ultrasound examinations by the Ministry of Health in Vietnam; three examinations or less and four examinations or more. The dependent variables with fixed response alternatives are presented in box 1. For the statements related to ‘the role of ultrasound for clinical management’ and ‘resources and training of obstetric ultrasound’, the response options were dichotomised into disagree or strongly disagree and agree or strongly agree in logistic regression analyses. The question ‘do you have a role in decision-making regarding clinical management on the basis of obstetric ultrasound examinations’ was used both as an independent and dependent variable, and the response options were dichotomised into no and yes for some analyses. The response options for the statements related to ‘improving utilisation of ultrasound’ were categorised as not at all or not very much and a fair amount or a great deal in logistic regression analyses. The response option neutral or don’t know was not included in either of these categories. For most statements, the response options neutral and don’t know were selected by a small proportion of the participants. How do you rate your skills in ultrasound in relation to the assessment/evaluation of: *Response options: never, on a daily basis, on a weekly basis, on a monthly basis, more seldom than on a monthly basis.†Response options: no skills, low skill-level, intermediate skill-level, high skill-level.‡Response options: no, yes a minor role, yes a moderate role, yes a major role.§Response options: not at all, not very much, a fair amount, a great deal, don’t know.¶Response options: strongly agree, agree, neutral, disagree, strongly disagree. *Response options: never, on a daily basis, on a weekly basis, on a monthly basis, more seldom than on a monthly basis. *Response options: never, on a daily basis, on a weekly basis, on a monthly basis, more seldom than on a monthly basis. †Response options: no skills, low skill-level, intermediate skill-level, high skill-level. †Response options: no skills, low skill-level, intermediate skill-level, high skill-level. ‡Response options: no, yes a minor role, yes a moderate role, yes a major role. ‡Response options: no, yes a minor role, yes a moderate role, yes a major role. §Response options: not at all, not very much, a fair amount, a great deal, don’t know. §Response options: not at all, not very much, a fair amount, a great deal, don’t know. ¶Response options: strongly agree, agree, neutral, disagree, strongly disagree. ¶Response options: strongly agree, agree, neutral, disagree, strongly disagree. For categorical variables, frequencies and percentages were analysed and Pearson’s χ2 test was used for test of difference, with the level of significance set at p<0.05. For continuous variables, mean values and their SDs were presented. Univariate and multivariable logistic regression was undertaken and presented with ORs and their 95% CIs. The independent and dependent variables used for logistic regression are reported in their specific sections as well as in box 1. All independent variables were entered into the univariate logistic regression analysis; however, only the statistically significant variables were included in the final multivariable logistic regression models. Statistical analyses were performed using SPSS V.23.