the Northern region of the Netherlands — the organization of tasks and responsibilities regarding risk assessment after the booking appointment is varied or unclear. We found no significant difference between MRS and professionals’ level of satisfaction. Based on these findings, we recommend an evidence- and value-based approach in the pursuit of the optimal organization of risk selection. The effectiveness of the practice of risk selection In two of our studies, we examined the effectiveness of the practice of risk selection. We studied the association betweenmaternal characteristics and perinatal mortality in healthy pregnancies as these associations are used as a rationale for offering induction of labour. We used descriptive and inter-categorical regression analyses to examine the association between single and multiple maternal characteristics ― maternal age, maternal parity, ethnicity and socioeconomic status (SES), and gestational age ― and the probability of fetal and neonatal mortality. The results showed that the probability of perinatal mortality differed among subgroups of women when taking multiple maternal characteristics into account. We observed that the highest probability of perinatal mortality was among nonDutch, low-SES women, nulliparous women aged ≥ 40 years, and women aged up to 29 years with ≥ 3 previous births. The findings of this study indicate that decision making about the use of induction of labour to prevent fetal and neonatal mortality based on a single determinant may lead to the overuse or underuse of induction of labour. A valuebased health care strategy, addressing social inequity, and investing in better screening and diagnostic methods that employ an individualized and multi-determinant approach, may be more effective in preventing fetal and neonatal mortality. We also studied variation in protocols as a cause of ineffective risk selection. We analysed the protocols used for neonatal referral in Dutch hospitals, looking at variation in the content of obstetrics and neonatal protocols for six common indications for neonatal referral: large for gestational age/macrosomia, small for gestational age/fetal growth restriction, meconium stained amniotic fluid, vacuum extraction, forceps extraction, and caesarean section. We compared the recommendations for care in the protocols and examined the association between that variation and a hospital’s geographical location and type of department. We identified large variation in recommendations for referral, admission, screening/diagnostic tests, treatment, and discharge. Our analyses revealed considerable inter-hospital, inter-department, and intra-department variation in recommendations for type of referral, admission, screening/diagnostic tests, treatment, and discharge. Furthermore, we found a lower referral threshold — leading to higher referral rates — in neonatal departments compared to obstetrics departments, and a higher referral threshold — leading to lower referral rates — in the Eastern region compared to the weighted referral threshold of all regions. To reduce unwarranted variation between local protocols, we recommend investing in the development of robust 161 7 GENERAL DISCUSSION