132 6 CHAPTER 6 4 | GOODARZI et al. We compared the three referral threshold categories “low threshold”, “average threshold”, and “high threshold” using dummy variables: “low threshold versus average and high threshold”, “average threshold versus low and high threshold”, and “high threshold versus low and average threshold”. We used the obstetric department as reference category for type of department. For region, instead of using one of the regions as normative category, we used the weighted referral threshold of all regions as the reference category. This weighted referral threshold was the overall referral threshold, with the referral threshold of the region weighted for the number of responding departments in each region. 3 | RESULTS 3.1 | Characteristics of the study sample Among the 74 hospitals that were approached, nine hospitals (12%) were tertiary-level hospitals. We received protocols from 83 departments: 39 (53%) obstetric departments and 44 (59%) neonatal departments. Of these departments, 22 (30%) were located in the same hospital. Nine (12%) hospitals did not respond to our invitation. Four obstetric departments (5%) and two neonatal departments (3%) declined participation—one obstetric department because of time constraints, and one neonatal department because their protocols were being revised at the time of this study (Figure 1, Table 1). 3.2 | Interdepartmental variation: between obstetric and neonatal departments located in different hospitals From the 83 responding departments, we received 420 protocols, together containing 68 indications for neonatal referral to the pediatrician. We found the highest number of different indications for neonatal referral in the SGA protocols and in the protocols received from the neonatal departments (Table 2, a detailed overview of the indications for referral can be found in Supplement 2). In most protocols, the date of publication, type of collaboration, and references were not mentioned. Of the protocols in which this information was given, the mean date of publication was 2016 and 2018 for neonatal departments and obstetric departments, respectively. Of these, most were hospital-only protocols that did not apply to the whole region, developed using a multidisciplinary approach, and based on guidelines. None of the protocols contained information about patient involvement in protocol development (Table 3). Most of the protocols did not contain a specific recommendation about location of admission and type of referral. If type of referral was specified, this concerned a pediatric consultation. Sometimes, the recommendation was not to refer to the pediatrician. Pediatricians' attendance at birth was just as often indicated as not indicated (21%) but most often was not specified (51%) (Table 4). In many of the protocols, the indications for referral were specified for different subindications. For F IGURE 1 Response rate specified for type of department 1523536x, 0, Downloaded from by Vrije Universiteit Amsterdam, Wiley Online Library on [14/11/2022]. See the Terms and Conditions ( on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License