Background

In this study, we investigate the decision-making practices of public health nurses (PHNs) operating in primary care settings at Norwegian Child and Family Health Clinics (CFHC), particularly regarding their approach to identifying and legally reporting suspected cases of child maltreatment. Given the pivotal role these professionals play in maintaining ongoing interactions with children and families, CFHC practitioners, including PHNs, are strategically positioned to recognize and respond to instances of child maltreatment (Gilbert et al., 2009).

Child maltreatment, also known as child abuse and neglect, refers to actions or behaviors that harm children or fail to protect them from harm. The World Health Organization (WHO) defines child maltreatment as “all forms of physical and/or emotional ill-treatment, sexual abuse, neglect or negligent treatment or commercial or other exploitation, resulting in actual or potential harm to the child’s health, survival, development or dignity in the context of a relationship of responsibility, trust or power” (WHO, 2020). Child maltreatment represents a severe global problem, carrying profound implications for children, families, and society as a whole (Graves & Gay, 2022; Irigaray et al., 2013; Schols et al., 2019; Segal et al., 2021; Strathearn et al., 2020). A meta-analysis indicates a notable rise in both health-related and economic burdens attributed to child maltreatment such as adverse childhood experiences throughout Europe (Hughes et al., 2017). The prevalence of child maltreatment exhibits variability across studies. In Stoltenborgh et al.’s (2015) review of a series of meta-analyses, it was found that the overall estimated prevalence for self-reported studies was 13% for sexual abuse, 23% for physical abuse, 26% for emotional abuse, and 24% for neglect (Stoltenborgh et al., 2015). These numbers are similar to a Norwegian report involving adolescents 12–16 years, where 20% reported being victims of physical abuse, with 5% describing serious abuse, “…such as being beaten up, beaten with an object, or beaten with a fist” (Hafstad et al., 2020). The occurrence of different types of child maltreatment varies depending on the child’s age. Specifically, sexual abuse is more common during adolescence, while physical abuse poses a greater risk for infants, often resulting in serious consequences (U.S. Department of Health and Human Services [HHS], 2022). Furthermore, while believed to be underreported, a significant number of infant fatalities can be linked to severe child abuse and neglect (Palusci et al., 2019). Despite its global significance, instances of child maltreatment often go unrecognized or unreported, leading to delayed interventions and prolonged suffering for affected children (Sethi et al., 2018).

In many countries, laws require individuals to report any suspicions of child maltreatment (Mathews, 2015). Health professionals, in particular, have a crucial role in preventing child abuse by regularly evaluating children, identifying risk factors within the family or environment, and reporting any suspected maltreatment to the appropriate authorities (WHO, 2014). Despite obligations to report, previous research indicates that recognizing and measuring childhood maltreatment can be difficult (Naughton et al., 2018; Vollmer‐Sandholm et al., 2021).

Child and Family Health Clinics (CFHCs) play a crucial role in Norway’s healthcare system, offering universal services through a standardized program. This program includes a minimum of 14 check-ups before a child reaches school age, with most visits occurring in the first year of life (Norwegian Directorate of Health, 2019). PHNs, who are registered nurses with additional training in health promotion and preventive care, serve as primary caregivers at CFHCs. They specialize in the psychological, physical, and emotional development of children. Given that almost all children and families in Norway (99%) attend their scheduled visits at CFHCs (Statistics Norway, 2022), PHNs are uniquely positioned to identify and prevent child maltreatment through regular engagement with families during early childhood.

The legal obligation for PHNs to report suspected child maltreatment is outlined in the Child Welfare Act § 13–4 and further specified in the Health Personnel Act §33. According to these laws, PHNs must report if they have reasonable suspicion that a child is being abused or neglected, based on specific circumstances related to the child’s situation. The laws emphasize that the reporting obligation doesn’t require absolute certainty, but rather a careful evaluation of the child’s circumstances.

However, a government report (NOU, 2017, p. 12) examining cases of child maltreatment in Norway found that many situations requiring intervention were missed, resulting in delayed responses. In 2019, the National Criminal Investigation Service in Norway (KRIPOS) published a report on severe child maltreatment cases involving children aged 0–4 years. The report found that infants less than a year old were involved in over 70% of the cases, with half of these cases involving children younger than 5 months old (KRIPOS, 2019).

Previous research within primary care has indicated that health professionals working in primary care acknowledge the importance of detecting child maltreatment, but also highlight the difficulties and complexity involved (Alfandari et al., 2022; Dahlbo et al., 2017; Midtsund et al., 2023; Nouman et al., 2020). In a recent qualitative Norwegian study, PHNs experienced insecurity when reporting maltreatment to CPS as they were uncertain whether their assessments were serious enough and had concerns about what would happen to the child if they filed a report (Midtsund et al., 2023). A mixed-method study among 59 Swedish nurses in primary care revealed a discrepancy between their belief in the importance of assisting families and their actual capacity to provide such help (Engstrom et al., 2021). They had little formal training and education on risk factors, lacked structured methods to address them, and were unclear on where to refer parents for help and which resources were available (Engstrom et al., 2021). Furthermore, a meta-synthesis of qualitative studies discovered that health professionals face difficulties in detecting and addressing more subtle forms of maltreatment, such as cases where there are no clear physical signs like bruises or injuries, and the symptoms are ambiguous and hard to specify (McTavish et al., 2017). According to a systematic review conducted by Wilson and Lee (2021), there are multiple factors that can hinder the reporting of incidents. These barriers can stem from structural limitations, resource limitations, insufficient support systems, sociocultural influences, and individual characteristics (Wilson & Lee, 2021). Regrettably, there exists no universally accepted gold standard or diagnostic test for identifying instances of child maltreatment, thereby presenting challenges in detection. The Norwegian guidelines for PHNs recommend that decision-making processes include integrating diverse informational cues and considering all the information as a whole to assess whether it constitutes child maltreatment (Norwegian Directorate of Health, 2019). Additionally, the cumulative effect of all information is crucial in determining the course of action. Consequently, the determination to report suspected cases hinges upon the subjective judgments, knowledge base, and professional dispositions of PHNs. Previous studies within primary care highlight child maltreatment decisions must be made in contexts characterized by uncertainty and unpredictability, and the practitioners are required to swiftly make difficult and emotionally charged decisions despite having limited information available (Enosh et al., 2021; Nouman & Alfandari, 2020).

According to Enosh et al. (2021), a vital approach to reducing biases in complex decision-making involves ensuring thorough oversight and critical evaluation of a professional’s own assumptions, reasoning, and practices. When reviewing the extant literature pertaining to the identification and response to child maltreatment, the majority of studies were conducted within hospital-based settings involving healthcare professionals by using vignette studies (Alfandari & Taylor, 2023; Naughton et al., 2018; Otterman et al., 2017; Panagopoulou et al., 2023; Vollmer‐Sandholm et al. 2021). Vignettes are fictional scenarios and are sought to be a valuable tool in surveys concerning suspected child maltreatment situations, as they enable the exploration of complex and sensitive issues (Bradbury-Jones et al., 2014). In early 1990, Misener (1986) and Zellman (1990) illustrated the usefulness of vignettes in understanding the severity of child maltreatment and the decision-making process of mandated reporters (Misener, 1986; Zellman, 1990). A recent vignette study conducted among Norwegian pediatricians in hospital settings revealed significant divergence and a lack of consensus in clinical decision-making processes, as well as inconsistency in adhering to national guidelines. Furthermore, the study underscored apparent gaps in knowledge and substantial disparities in the management of suspected cases of child maltreatment (Vollmer‐Sandholm et al., 2021). However, hospitals operate under different conditions for assessment than PHNs in CFHC, as they likely encounter a higher volume of children with injuries compared to primary care settings, where appointments are scheduled rather than emergent. PHNs may be the most consistent health professionals in children’s early life, as they follow them over time from birth until school age. Understanding how PHNs perceive and handle cases of suspected child maltreatment is essential for developing effective intervention strategies and support systems. There is limited research on how health professionals in primary care perceive and respond to possible child maltreatment situations, and to our knowledge, no quantitative study has specifically examined PHNs in Norway.

This study aims to investigate PHNs’ self-perceived competence and management in cases of suspected child maltreatment and investigate how judgments are made.

Method

Design and Sample

This study adopts a one-phase convergent mixed-method design according to Creswell and Plano (2018), where both qualitative and quantitative data were collected concurrently. A cross-sectional survey, incorporating closed-ended and open-ended questions, was administered to PHNs working within CFHC between October 24 and December 31, 2022. Through the utilization of a patient vignette questionnaire, the study assessed PHNs’ competence and clinical decision-making processes. The subsequent analysis of the qualitative and quantitative data was done separately. This integrated approach, as advocated by Creswell (2023), converges both quantitative and qualitative data to facilitate a comprehensive analysis of the research problem. This approach enables a more complete understanding by comparing quantitative and qualitative results.

PHNs were recruited from the Norwegian Union of Public Health Nurses’ register, the only professional association for PHNs in Norway. An electronic questionnaire was sent to all members with a valid email address, amounting to 3798 of the 3896 members (97.5%). To encourage response rates, reminder emails were sent after 9 days, 3 weeks, and 7 weeks.

Eligible participants were PHNs who consulted with families and worked with children aged 0–5 years at CFHCs. As we did not have information on where in the service the PHNs worked, we first asked for demographic data. At question 7, PHNs not working at the CFHC 0–5 were excluded. Initially, 1873 PHNs responded to the survey. Among them, 1238 were excluded because they did not work in a CFHC, and 45 more at question 9 for not participating in family consultations. Ultimately, the survey was completed by 554 PHNs, and the final analysis incorporated these 554 responses.

Ethical Considerations

Approval was obtained from the Norwegian Service for Shared Service in Education and Research (SIKT), registration number 303782. Participants received the study information sheet and survey link. Informed consent was signified by completing the questionnaire; no identifiable information was gathered. The excluded participants were appropriately informed of their ineligibility and given the chance to withdraw their demographic data. PHNs not included in the study were not asked about child maltreatment. The study adhered to the Declaration of Helsinki II recommendations (World Medical Association Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects, 2013) (World Medical Assembly, 2013).

Instrument

A customized survey instrument was developed for this study as part of the project titled “The Role of Public Health Nurses in Child and Family Health Clinics in Preventing and Detecting Child Maltreatment.” The questionnaire development process, detailed in a previous study (Midtsund et al. 2024), comprised five sections: demographics, guidelines and experience, vignettes, knowledge, and attitude (Supplementary file1). This article will focus on the vignettes and associated questions about PHNs’ level of concern for child maltreatment, the reason for the concern, self-perceived competence, and which clinical management, reporting, and follow-up actions would be taken.

The survey utilized three potential maltreated children vignettes. Vignettes one and two in the survey were created in collaboration with a reference group comprising six PHNs with diverse backgrounds and experiences, including a CFHC manager, a PHN lecturer, three practicing PHNs, and a recent PHN graduate. This group ensured the scenarios were realistic and aligned with the responsibilities of PHNs. Vignette number three was adapted, with permission, from another questionnaire (Vollmer‐Sandholm et al. 2021), and further modified to suit a CFHC context. The three vignettes necessitated an examination of potential physical, sexual, and emotional/psychological maltreatment and neglect in children aged 4 years, 15 months, and 3 months (see Fig. 1 for vignette summary and Supplementary file 2 for full text). The first question following each vignette was “How much do you agree/disagree with the following statement: I am concerned that this child has been subjected to child maltreatment, and it is necessary to investigate further.” Similar to Vollmer‐Sandholm et al. (2021), we used a 7-point Likert scale response ranging from “strongly agree” to “strongly disagree.” Likert scale, particularly the 7-point version, is widely used for evaluating levels of agreement/disagreement due to its enhanced sensitivity (Taherdoost, 2019).

Fig. 1
figure 1

Summary of vignettes

Full size image

Subsequently, an open-ended question was posed to elucidate the reason for their concern. To prevent directing the PHN towards a specific viewpoint, the follow-up question was contingent on their initial response. For those who responded with “strongly agree” to “ partially agree,” the follow-up question was alternative 1: “Describe your concerns.” Conversely, for those who responded with “Neither agree nor disagree” to “strongly disagree”, the follow-up question was alternative 2: “Any comments?” Both questions included text response options, offering participants the opportunity to provide supplementary information regarding their answers.

The next question in the vignette section aimed to assess self-perceived competence in each described vignette through the statement: “I feel confident that I have the competence and knowledge to handle this case.” Those who responded with “Neither agree nor disagree” to “strongly disagree” were then prompted to comment on their response.

Finally, we established clinical management options in each case. In Vollmer‐Sandholm et al.’s (2021) questionnaire, clinical management included medical investigations (e.g., blood tests and X-rays), which were not applicable in the CFHC context. Therefore, we adjusted the management alternatives to suit our population, following the recommendations from the clinical guidelines (Norwegian Directorate of Health, 2019). However, options such as reporting to CPS, reporting to the police, follow-up, and discussion with colleagues remained consistent across both surveys. We found that most questions had intended responses for each case.

Data Analysis

Quantitative Analysis

The analysis of quantitative data was performed using IBM SPSS Statistics software, Version 28.01.01. The 7-point Likert scale used where recorded and 1 corresponds to “strongly disagree” and 7 corresponds to “strongly agree.” The summary of the data is presented as frequency, percentage, and mean.

Qualitative Analysis

The qualitative content analysis, inspired by Graneheim and Lundman (2004), guided the analysis of the textual responses from the collected data, focusing on the two successive follow-up inquiries (Graneheim & Lundman, 2004). Manifest content analysis examined the explicit text to develop sub-categories and categories, while latent content analysis interpreted underlying meanings to identify a theme. The overarching objective was to perform a systematic and reliable representation of the text. Consistent with the study’s objectives, written responses from each vignette were brought together into one text document, constituting the unit of analysis.

The initial phase of analysis entailed identifying meaning units closely aligned with the text, followed by condensing the text into shorter units for subsequent coding. The meaning units are words, sentences, or paragraphs containing aspects related to each other, or to their content and context. Meaning units were assessed relative to the aim of the study, which illuminates the trustworthiness of the process (Elo et al., 2014). Because the responses in the survey instruments comprised brief sentences, extensive text condensation was unnecessary, facilitating coding into words or shorter phrases. Next, descriptions were abstracted into codes, subcategories, and categories across varying levels, as described in Table 1. These codes retained the essence of the data and were classified under 9 subcategories for Vignette 1, 8 subcategories for Vignette 2, and 10 subcategories for Vignette 3. Finally, the subcategories extracted from the diverse vignettes were subsequently organized and grouped into three comprehensive categories encompassing all three vignettes.

Table 1 Example of analytic meaning units, condensed meaning units, code, subcategory, and category (drawn from Vignette 3)
Full size table

To finalize the analytical process, we integrated the core insights from the subcategories to develop one overarching theme. Ensuring the credibility of our analysis required collaborative efforts from all four authors. Initially, the first and last authors independently reviewed the initial ten pages of text, identifying key units of meaning and corresponding codes concurrently. Following this, the entire analysis process underwent collective scrutiny, discussion, and reflection by all authors to uphold thoroughness and reliability.

The participants who responded strongly disagree to neither disagree nor agree were given the opportunity to provide additional comments in response to the open-ended follow-up question; “any comments.” These comments collectively amounted to three pages of text. Although all comments were thoroughly read and discussed, they were not uploaded into Nvivo due to the limited amount of text available.

In the second question following the vignettes about self-perceived competence, PHNs who selected alternatives 1–4 (indicating disagreement to neutrality) were prompted with a text box to provide voluntary comments. The collective volume encompassed a total of 64 comments across the three distinct vignettes, and the text comments were imported into NVivo and systematically coded. In this analytical endeavor, the initial focus was directed towards identifying codes manifested within the textual content, subsequently discerning patterns of recurrence in the responses.

The results of both the quantitative and qualitative responses will be initially presented separately, followed by their integration to offer a comprehensive understanding of PHNs’ comprehension, reasoning, and practice. Additionally, other qualitative follow-up questions prompted by specific survey inquiries, such as “specify” or “comments,” will be interspersed throughout the text.

Result

Demographics

Five hundred and fifty-four PHNs responded to the questionnaire, and the majority of PHNs were aged between 31 and 50 years (Table 2). Most PHNs (97.3%) were educated as registered nurses with a PHN degree. Moreover, 39.2% of respondents held additional education or qualifications beyond a PHN degree. Subsequent inquiries into what supplemental education demonstrated that the majority had undertaken specialized training in diverse nursing disciplines, encompassing pediatric nursing, midwifery, intensive care nursing, psychiatry, and migration health. Furthermore, several respondents indicated having education in family therapy, lactation consultancy, administration/management, or mental health. Additionally, 87 PHNs noted that they possessed a master’s degree.

Table 2 Demographics of the PHN1 respondents (N = 554)
Full size table

Quantitative Responses to the Child Maltreatment Vignettes

Table 3 summarizes PHNs’ response to the two statements that followed the three vignettes regarding concerns about whether the child may have been subjected to child maltreatment and neglect and PHNs’ self-perceived confidence in conducting assessment in the described scenarios. For the first statement, the participants generally showed a high agreement about being concerned the child may have been subjected to child maltreatment across all cases, with Vignette 1 the 4-year-old disclosed being beaten and Vignette 3 the 3-month infant with bruises, exhibiting the highest agreement score (Fig. 2).

Table 3 PHN1 level of concerns and self-perceived confidence in child maltreatment vignettes
Full size table
Fig. 2
figure 2

Level of concerns

Full size image

Qualitative Analysis of Responses Regarding Concerns

Among the PHNs who expressed concerns about child maltreatment (ranging from partially agree to strongly agree in the quantitative section), 476 out of 516 (92%) provided textual responses of their concerns to Vignette 1 (V1). All 494 (100%) PHNs provided textual responses to Vignettes 2 (V2) and 512 (100%) in Vignette 3 (V3). The analysis generated a main theme (latent), presenting three overarching categories (manifest) across all vignettes, as depicted in Table 4. The main theme “A Holistic and Child-Centered Risk Appraisal” was generated from comments across all vignettes, reflecting three categories that underpinned the PHNs’ assessments and concerns of child maltreatment. These categories were “Individual Risk Assessments,” “Family Risk Assessments,” and “Complicated Risk Contemplations,” in which PHNs considered multiple aspects, including the child’s needs, safety, and well-being. The analysis demonstrated that the PHNs grounded their concerns of assessments in a comprehensive evaluation of both the individual child and the family unit. They provided holistic, reflective summaries that were tailored to the specific scenarios presented in the vignettes. Further, the findings are presented with categories and substantiated by citations from the text responses.

Table 4 Main theme, categories, and subcategories
Full size table

Individual Risk Assessments

Concerns based on individual risk assessments were prominently featured across all three vignettes. These assessments focused on various concerns related to the child’s physical and emotional well-being, including observable signs such as physical injuries and behavioral indicators. In V1 and V3, significant emphasis was placed on observable indicators, such as the child’s disclosure of being beaten, weight fluctuations, and restlessness. For instance, responses such as “the child says he’s being hit,” “the child has had a significant weight gain,” and “restless child” were noted in V1. Similarly, in V3, physical signs like bruises and a torn lip frenulum, as well as emotional distress, were key focal points. Examples include responses such as “such young children should not have bruises” and “the child is inconsolable.” In V2, concerns were predominantly related to the child’s emotional state and lack of social support, with comments such as “the child appears fearful,” “the child does not receive comfort,” and “it is concerning that the child does not attend daycare and has limited contact with others.”

Family Risk Assessments

Family risk assessments addressed the broader context of the child’s environment, focusing on family dynamics and support systems. V2, in particular, featured significant concerns about the family’s psychological state and interactions, with comments like “the mother appears distressed, and the father seems depressed as he just sits there” and “dissatisfied mother, how does she function as a mother?” highlighting parental vulnerabilities. A great number of concerning comments related to family risk assessments were mentioned in both V1 and V3. In V1, family burdens such as divorce and isolation due to a limited network were noted, along with remarks about the mother’s behavior, particularly her silence when the child disclosed being beaten. There were also comments regarding the mother’s mental health and the increased risk of stress due to the parental breakup. In V3, comments reflected concerns about parental stress and lack of regulation skills. The parents were described as vulnerable due to their youth, having a child with colic, and appearing stressed and exhausted, including their inability to comfort the child after vaccination, as indicated by responses such as “young parents, lack of comforting ability, only large rocking movements and fiddling with the toy” (V3).

Complicated Risk Contemplations

Complicated risk contemplations involved deeper reflections on the multifaceted nature of concerns and risks, including cultural and contextual factors. Many comments reflected an internal dialogue about the complexities of the scenarios, considering both the potential for maltreatment and the need for additional support. For example, in V1, comments such as “Although it may not necessarily involve severe violence, the mother may require help in managing her anger” reflected this nuanced consideration. In V2, reflections included assessing cultural and social norms, with comments like “They come from a different culture where corporal punishment is more common. They might hit the child when they’re frustrated.” In V3, responses included concerns about the plausibility of the explanations provided and requests for further information, such as “How did the child manage to do this? An explanation is needed here” and “I would like to know what happened at the emergency room. Have they filed a report there?”.

These qualitative responses substantiate and correlate with the quantitative data regarding PHNs’ level of concern and the necessity for further investigation to ensure child safety. They provide a detailed elucidation of the underlying foundations and bases for these concerns.

Finally, the textual responses from the PHNs, who across all the vignettes whose responses indicated that they were not concerned about the child being at risk of maltreatment but were prompted with the follow-up question “Any comments?” instead of “Describe your concerns.” Among these PHNs, 16 of 37 provided comments for V1, 26 out of 60 for V2, and 18 out of 42 for V3. These comments revealed that many were worried despite their initial response. Comments across all vignettes primarily highlighted the necessity for further investigation and assessment, such as “I would discuss my concerns with the mother and promptly schedule a follow-up consultation to delve deeper into the situation” (V1). PHNs expressed concerns for the well-being of the children and emphasized the importance of follow-up actions, scheduling follow-up consultations, and ensuring child safety. In contrast to the responses from concerned PHNs, the responses from PHNs who were not concerned revealed a discrepancy in the qualitative data. Their comments highlighted widespread concerns among respondents despite the level of concerns in their initial responses.

Qualitative Responses Regarding Self-Perceived Competence

Regarding PHNs’ self-perceived competence in the given scenarios, there was a slightly less agreement compared to their concerns. They still exhibited high confidence in their assessment competencies (Fig. 3). The subsequent textbox, which allowed respondents to add comments to their responses, was activated for 57 respondents in V1, 63 in V2, and 57 in V3. These respondents selected options 1–4, indicating a range from disagreement to neutrality in their self-perceived competence. Among these, 26 (25%) participants provided comments in V1, while 18 (28%) PHNs left comments in V2, and 20 (35%) respondents commented in V3. The textual analysis of these comments yielded two discernible categories “thoughts immersed with uncertainty” and “contemplations on important actions.”

Fig. 3
figure 3

Level of self-confidence

Full size image

The majority of the written text leaned towards the category of “Thoughts immersed with uncertainty,” where most comments were questions seeking additional information about the situation presented in the vignette. Many wrote that the case was difficult to assess, and they were uncertain whether the symptoms described were enough to qualify as maltreatment. For instance, one respondent noted, “I am unsure if what has been observed is ‘enough’” (V3). Some wrote that they had knowledge about maltreatment but little experience in practical situations, which made them insecure, as exemplified by one respondent: “I have knowledge about violence and abuse, but I don’t feel confident enough in my expertise to address this at that moment” (V1). Others articulated a desire for professional guidance in navigating such complex scenarios and wanted to seek counsel from peers or relevant authorities. Additionally, several commented that they possessed the theoretical knowledge and competence about maltreatment but needed more information about when and what would be appropriate action. Some comments were about specific actions, such as “The child needs medical supervision/evaluation” (V3) or “I would contact the family counseling office” (V2). The textual responses confirm the quantitative responses regarding competence in the vignettes, as it was the respondents who expressed disagreement to neutrality in self-competence who had the opportunity to provide textual responses. The data from these textual responses suggested a high degree of uncertainty, providing additional context to the self-confidence indicated in the initial quantitative response.

Managing Child Maltreatment Vignettes

Table 5 presents the actions that PHNs would undertake in the three distinct scenarios. Generally, responses varied in how they would handle the situations, with actions recommended in the guidelines, such as assessing the child, exploring the situation, and documenting, being more frequently repeated across the three vignettes. Additionally, other commonly reported actions included scheduling a follow-up appointment and discussing the situation with colleagues. Fewer PHNs indicated they would refer the child for examination at a hospital or by a GP, and reporting to the police or CPS was low across all vignettes, with a total of 26.9% in V1, 10.7% in V2, and 51.5% in V3.

Table 5 PHNs’ management of suspected child maltreatment vignettes
Full size table

PHNs who selected “other measures” were given the opportunity to leave a comment with suggestions for other actions they wished to undertake. The written comments were reviewed and organized.

In all three vignettes, comments primarily focused on analyzing the situation and discussing potential actions. Suggestions included referring families to programs such as organized child maltreatment prevention programs, family counseling services, mental health services, and other childcare interventions. Some proposed referring the whole family to a residential family assessment center. Overall, the comments reflected more PHNs’ desire for better understanding and confidence in their actions than additional suggestions of measures as exemplified by this text response from V1: “I would investigate a bit more before reporting to CPS, as what the child says might not be true. However, the overall assessment is that this is a family that might need some form of follow-up for many reasons. Is the child sick? What resources do they have around them? How is the cooperation between the mother and father, etc.?”.

Discussion

This study aimed to investigate PHNs’ self-perceived competence and management strategies in cases of suspected child maltreatment, as well as to investigate the processes by which they make their judgments. The majority of PHNs identified child maltreatment in the vignettes and displayed a relatively unanimous level of concern. Qualitative text responses provided further insights into their emphasis on concerns of child maltreatment, aligning with the quantitative responses.

The qualitative findings underscored the complexity and uncertainty inherent in the assessment process, a phenomenon well-documented in prior research (Nouman & Alfandari, 2020; Wilson & Lee, 2021). Similar to Nouman et al.’s findings among healthcare professionals in community health settings, the assessment process was not linear but comprised several steps, including recognition of signs of child maltreatment, preliminary interpretation, and justifications for one’s judgment (Nouman & Alfandari, 2020). In our study, uncertainty predominantly centered around self-confidence and the adequacy of one’s judgments, particularly regarding the threshold for reporting. This uncertainty was reflected in their management approaches, with the most supported measures focusing on further exploration of the situation and strengthening confidence in their assessments, rather than reporting to CPS, despite having concerns about potential child maltreatment.

Consistent with previous studies (Enosh et al., 2021; Hoffmann Merrild et al., 2023; Visscher & Van Stel, 2017; Vollmer‐Sandholm et al. 2021), many PHNs in our study responded they would discuss their concerns with other colleagues. Nouman et al. (2020) found that decision-making in primary care was enhanced by reflecting on perceptions and assessments with other professionals. The significance of colleagues in the decision to report is earlier described by Tufford and Lee’s (2019) vignette study among social workers, where one of the top three decision-making factors for study respondents when reporting suspected child maltreatment was the opinion of your colleague(s). However, the work structure at Norwegian CFHCs involves PHNs conducting individual consultations independently, which contrasts with the multi-professional approach found in hospitals. This structural difference may limit opportunities for such discussions. Additionally, varying staffing levels and time constraints at CFHC, as illustrated in V1, by the next child in PHNs timetable already waiting in the reception area, further restricts access to collaborative dialogue. Consequently, it remains unclear whether the expressed action translates into practical facilitation within CFHC settings or merely reflects a desire to engage in such practices. Lack of time as a barrier is well-documented in previous research about work against child maltreatment (Adams et al. 2022, Dahlbo et al. 2017, Midtsund et al. 2023). In a recent qualitative study, Norwegian PHNs indicated that the time-intensive nature of addressing child maltreatment is particularly challenging due to the substantial demands of their numerous mandatory tasks (Midtsund et al. 2023).

The discrepancy between high recognition of concerns in hypothetical maltreatment scenarios and low reporting to CPS and police is evident and consistent with prior research (Sathiadas et al., 2018; Vollmer‐Sandholm et al. 2021; Zusman & Saporta-Sorozon, 2022). However, in stark contrast to our findings, Vollmer et al.’s hospital study among 157 pediatricians found that less than 30% would plan follow-up consultations in the vignettes (Vollmer‐Sandholm et al., 2021). The Norwegian clinical guidelines for PHNs emphasize individualized follow-up in CHFC (Norwegian Directorate of Health, 2019), enabling PHNs to offer home visitations and more continuous care than hospitals, which typically focus on specific patient visits rather than universal monitoring. Further, the high response regarding further follow-up in our study may be explained in PHNs’ need for gaining more insight and thus corroborates the uncertainty expressed in the qualitative responses. The results of a mixed-methods study conducted in Danish primary care settings, involving the participation of 1252 general practitioners (GPs), suggest that determining severity based on a single consultation is challenging, with one strategy being scheduling follow-up consultations to monitor the child’s progress (Hoffmann Merrild et al., 2023).

Hoffmann Merrild et al. (2023) further argue that the uncertainty surrounding child maltreatment cases might be worsened by the limited experience healthcare professionals have dealing with such cases in primary care. Although PHNs encounter nearly all children, the chances of extensive practice in handling suspicions of child maltreatment will be low for most PHNs. As demonstrated in a cross-sectional study that examined PHNs’ clinical experiences with child maltreatment, the findings indicated that suspicion of all forms of maltreatment was rare, and the majority reported encountering fewer than two cases a year (Midtsund et. al 2024). The influence of clinical exposure and experience on reporting is demonstrated in the study by Vollmer‐Sandholm et al. (2021), which found that individuals employed in specialized units within pediatric departments, who routinely evaluate children for maltreatment concerns, reported a greater number of cases compared to their counterparts in general medical units. Other studies emphasize the influence of training on the effectiveness of reporting child maltreatment, as demonstrated in a large cross-sectional Greek hospital study where, despite the generally low reporting rate among participants, the majority of those who reported instances of child maltreatment had received training (Panagopoulou et al., 2023). In fact, training was one of the main findings of increased reporting in a recent scoping review on health professionals identifying and reporting child physical abuse (Bragança-Souza et al., 2024).

PHNs’ commitment to provide extra follow-up mirrors the findings in another study we conducted where PHNs reported extensive use of extra consultations and contact with families outside working hours (Midtsund et al. 2024). The qualitative responses in this study further revealed a variety of proposed interventions to ameliorate the situation for the family. In Hoffmann Merrild et al.’s (2023) study on primary care, it is posited that despite instances of underreporting, health professionals may be attempting to improve the child’s situation by leveraging the established relationship, providing follow-up care, and managing their uncertainty by involving others, working on relations with the family, or adopting a watchful waiting approach (Hoffmann Merrild et al., 2023). In social science, this process can be understood as a negotiation between the prevailing uncertainty and the need for control (ibid). Concurrently, their actions could be interpreted as an avoidance of reporting. In a recent qualitative study, one reason PHNs delayed reporting to CPS was due to uncertainty about the potential outcomes of their reports (Midtsund et al., 2023). Uncertainty has been well-documented in previous research (Alfandari et al., 2022; Enosh et al., 2021; Wilson & Lee, 2021). Other studies indicate that uncertainty is compounded by fear of the potential consequences for the child and the family if a report is made, including fear of the situation deteriorating rather than improving (Visscher & Van Stel, 2017). Given the findings of our study and previous research, the lack of ability to identify child maltreatment doesn’t seem to be the challenge, but rather the determination to report where fear and uncertainty are intertwined. The wording in the clinical guidelines for PHNs does not impose stringent requirements on when to report; the decision relies on subjective judgments, knowledge base, and professional dispositions. In an American descriptive statistic study investigating the impact of statutory wording on the frequency and manner of child maltreatment reporting across various states, researchers found that the legislative phrasing indeed influenced reporting patterns (Piersiak et al., 2023). For instance, states employing the term “suspicion” rather than “belief” in their laws had a higher number of potential child maltreatment reports, yet fewer of these reports were substantiated as actual abuse. Norwegian PHNs have called for more explicit guidelines to aid in the challenges they meet in detecting child maltreatment (Midtsund et al., 2023). The TEN-4-FACESp rule, proposed by Pierce et al. (2021), offers a valuable tool in this context. This rule considers bruises on the torso, ear, neck (TEN), frenulum, angle of jaw, cheeks, eyelids, subconjunctivae (FACES), and any patterned (p) bruises. The “4” signifies any bruising on an infant aged 4.99 months or younger. This rule is applicable only to children under 4 years old with bruising, and a positive response for any of these components signals a classification of abuse. Based on the findings from the screening of 21,123 children for bruising, a classification tree was developed and validated to distinguish between abusive and non-abusive trauma (ibid).

Previous research further indicates that even though health professionals suspect child maltreatment and do not report it immediately, the suspicion still often leads to seeking advice or reporting the concern later on (Enosh et al., 2021). This suggests that we may not always assume that healthcare professionals consistently recognize and report child abuse in the same manner, highlighting the importance of examining the specifics of their practices. At the same time, it is deeply concerning that only 52% of PHNs in our study indicated they would report to the police or CPS concerning the 3-month-old baby (V3). This illustrates a need for more concrete criteria and reporting thresholds to reduce ambiguity and facilitate informed decision-making by PHNs.

Strengths and Limitations

This study presents several notable strengths, particularly its utilization of a sizable sample from diverse geographical regions in Norway. Nonetheless, the absence of comprehensive data regarding the total population of PHNs in this context poses inherent limitations on the generalizability of the study’s findings to all PHNs working within similar primary care settings. Therefore, caution is warranted when extrapolating the study’s results to broader PHN populations.

The study is also subject to various limitations. Evaluating cases of maltreatment and neglect, even when presented in vignette form, is inherently challenging due to the subjective nature of interpretation and the varied contexts in which PHNs operate. Many questions remain unanswered when responding to such scenarios. Still, this inherent complexity can mirror real-life scenarios and underscore the critical importance of adhering to guidelines despite the subjective nuances present. The absence of a definitive gold-standard threshold for determining when to report further adds a layer of complexity to the analysis and interpretation of results, particularly because the work of child maltreatment is intricate and not straightforward. Moreover, the study’s advertisement specifically addressing the topic of child maltreatment may attract respondents with a heightened interest or awareness of the subject matter, potentially leading to a biased sample selection. This bias could skew the representation of respondents, thereby limiting the study’s overall representativeness and introducing inherent biases in the findings.

Implications

The findings of this study present several implications for addressing child maltreatment. Firstly, recognizing the inherent uncertainty in responding to suspicions of child maltreatment is crucial. By acknowledging this uncertainty, we can better support PHNs through targeted training programs. These programs should aim to enhance their confidence and equip them with practical strategies to navigate the complexities and uncertainties inherent in the assessment process.

Secondly, fostering interprofessional collaboration among healthcare professionals is paramount. Facilitating discussions among colleagues can not only enhance decision-making processes concerning child maltreatment reporting but also ensure a more efficient flow of information between services. This could contribute to greater confidence and predictability regarding the outcomes of reporting.

Additionally, a revision of the clinical guidelines may be beneficial to provide more explicit and precise directives concerning when to report suspected child maltreatment, such as the TEN-4-FACESp rule as mentioned earlier (Pierce et al., 2021).

Lastly, further research and evaluation are needed to assess the efficacy of these proposed interventions, including training programs, interprofessional collaboration initiatives, and revised clinical guidelines, in improving the reporting of child maltreatment.

Conclusion

The findings suggest that while PHNs are skilled at identifying potential child maltreatment cases, there is a gap in their confidence to handle these situations effectively. This discrepancy may stem from a lack of practical experience or uncertainty about the appropriate actions to take. Therefore, it is crucial to provide PHNs with additional training and support, focusing on practical strategies for handling suspected child maltreatment cases. This will possibly enhance their professional competence and may also ensure a more effective response to safeguarding children’s welfare.