Staff Responses When Parents Hit Children in a Hospital Setting

Abstract

Objective

Physical punishment of children is a prevalent practice that is condemned by most medical professionals given its link with increased risk of child physical abuse and other adverse child outcomes. This study examined the prevalence of parent-to-child hitting in medical settings and the intervention behaviors of staff who witness it.

Method

Staff at a children's medical center and a general medical center completed a voluntary, anonymous survey. We used descriptive statistics to examine differences in the experiences of physicians, nurses, and other medical staff. We used logistic regression to predict intervention behaviors among staff who witnessed parent-to-child hitting.

Results

Of the hospital staff who completed the survey (N=2,863), we found that 50% of physicians, 24% of nurses, 27% of other direct care staff, and 17% of non-direct care staff witnessed parent-to-child hitting at their medical center in the past year. A majority of physicians, nurses and other direct care staff reported intervening sometimes or always. Non-direct care staff rarely intervened. Believing staff have the responsibility to intervene and having comfortable strategies with which to intervene were strongly predictive of intervention behavior. Staff who did not intervene commonly reported that they did not know how to respond.

Conclusion

Many medical center staff witness parent-to-child hitting. Although some of the staff reported that they intervened when they witnessed this behavior, the findings indicate that staff may need training to identify when and how they should respond.

Physical punishment, including spanking or other forms of hitting a child, is associated with a host of negative outcomes for children, such as behavioral and mental health problems in childhood^1,2 and physical and mental health problems in adulthood.^3,4 The abundance of research linking negative child outcomes with spanking has led several professional organizations for health care workers, including the American Academy of Pediatrics and the National Association of Pediatric Nurse Practitioners, to take public positions advising against the use of physical punishment and for the use of non-violent alternatives.^5–7 Yet parent-to-child hitting enjoys widespread public acceptance in the U.S.: in 2014, a nationally representative survey found that 76% of men and 65% of women agreed that children sometimes need a “good hard spanking”.⁸ Furthermore, despite the fact that many view physical punishment less favorably than they do other disciplinary practices⁹, a majority of parents have used physical punishment on at least one occasion.¹⁰ Health care professionals, especially pediatricians, are trusted sources of parenting information,¹¹ and thus these professionals have the potential to influence parents' disciplinary beliefs and behaviors.^11–13

There are no previously published estimates of the prevalence of parent-to-child hitting in medical centers. However, as a common parenting tactic⁸, it is probable that it sometimes occurs in medical centers. When parent-to-child hitting occurs in their presence, health care professionals and medical staff must decide whether and how they should take action. The decisions staff make in these circumstances are important: although staff may be reluctant to intervene in parenting, a failure to intervene may be construed as tacit approval. Moreover, in cases where parent-to-child hitting is so severe that it meets the standard of child abuse, a failure to act may also have legal implications for hospital staff. In all states, health care practitioners are mandated reporters of child maltreatment¹⁴; yet, most states, including the states from which our data were collected, have no reporting requirements for medical center employees who are not involved with the direct care, treatment, or research of persons.¹⁴

Little is known about how often parent-to-child hitting occurs in hospital settings or about the factors that influence staff intervention behavior. A related body of research that may inform staff intervention behaviors pertains to bystander interventions. This research, though typically focused on sexual assault intervention, has shown that subjective norms that support intervening, and the perception that peers support intervening, are predictive of intent to intervene and of actual intervention behavior.^15,16 Thus, we consider whether staff attitudes toward physical punishment, beliefs about the responsibility of medical personnel to intervene, and perceptions of peer attitudes about physical punishment each influence staff intervention behaviors. Bystander intervention research has also shown that bystanders who are confident that they know how to intervene successfully are more likely to do so.^16,17 Similarly, staff intervention behaviors may be influenced by whether they feel knowledgeable about how to intervene.

Despite the important roles that medical professionals have as sources of parenting information and as mandated reporters of child abuse, we are aware of no studies that have examined how often medical center employees witness parent-to-child hitting in medical settings or how they respond when they do witness it. In this study, we used primary data collected through staff surveys at two medical centers (one pediatric and one general) to explore the attitudes, experiences, and behaviors of medical center staff who witness parent-to-child hitting. We also compared attitudes and experiences by job type (physicians, nurses, other direct staff, and non-direct care staff). We aimed to address four research questions: (1) How often do staff witness parent-to-child hitting in the medical center?; (2) How do staff react when they witness parent-to-child hitting?; (3) What factors predict whether staff intervene when they witness parent-to-child hitting?; and (4) What do staff report as their reasons for not intervening?

Method

Procedure

Two medical centers participated in the study. One is a large general medical center in Wisconsin and the other is a children's medical center in Missouri. In 2014, all staff members at each medical center (approximately 6,600 at the general medical center and 6,000 at the children's medical center) were invited to participate in a voluntary, online survey. (At the general medical center, some staff members also completed the survey in person after a staff training.) The questionnaire was written at a high school reading level and took approximately 10 minutes to complete.

Survey participation was anonymous and not linked to staff identities in any way. Given that the survey was anonymous, participants were instructed that by moving from the introductory page, they were indicating their consent. Each medical center had protocols in place to ensure the integrity of the survey. The link to the online survey was provided to staff via their hospital-affiliated e-mail address for both medical centers. Thus, it is unlikely that the surveys were accessed by anyone not affiliated with the medical centers. The general medical center also removed the possibility of repeated responses by a single staff member by making the link received by each staff member accessible only one time.

At both medical centers, direct care staff members are trained during orientation on their roles as mandated reporters of child maltreatment, the availability of social work supports, the roles of supervisors, and the availability of security on site to assist with challenging family interactions. Study procedures were approved by all relevant institutional review boards.

Measures

Staff members were asked a series of questions created for this study. The first set of questions asked staff if they had witnessed a parent hitting a child in the medical center (“In the course of your work, how often do you see or hear a parent hitting a child?”). Staff were then asked, during the past 3 months when they saw or heard this, whether they had intervened in the past three months, and, if so, how often (always, sometimes, or never) and how they had intervened. The set of questions on intervention behaviors were limited to those who witnessed hitting in the past 3 months due to concern about accuracy of recall for events further in the past. Respondents who intervened at some point in the last 3 months were also asked whether they were comfortable intervening and whether they felt their intervention was effective. They also reported how they intervened (e.g., talking with the parent, telling a supervisor). Among those who witnessed but did not intervene, we asked staff to report their reasons for not intervening.

Participants also responded to several attitudinal and knowledge-based items on a 5-category response scale, ranging from strongly disagree to strongly agree. To assess support for staff intervening in parent-to-child hitting, staff were asked 5 items about the role medical center staff should play when a child is being hit (e.g., “Physicians, nurses, and other medical center staff have an obligation to intervene when children are being spanked or hit in the medical center buildings.”). These items were combined into a scale, with a respondent's score equal to the mean of their responses across the five items, and demonstrated strong internal reliability (α= .80).

Staff members were asked if they had comfortable strategies to intervene if they saw a parent hitting a child (1 item: “I have some comfortable strategies that I can use to intervene when I believe a parent is administering excessive or inappropriate discipline to a patient or sibling.”).

Staff members were asked about their own attitudes about spanking (9 items; e.g., “There are better ways to discipline a child than to spank them.”) using 5 items from the Attitudes toward Spanking scale¹⁸ and four additional items created for this survey.

Staff members also reported their perception of their coworkers' positive attitudes about spanking using 4 items from the Attitudes toward Spanking scale rewritten to ask about coworker attitudes (e.g., “Most of my co-workers think that spanking is a bad disciplinary technique.”).¹⁸ Each scale demonstrated strong internal reliability (respondents' spanking attitudes: α=.89; perceived coworker attitudes: α=.88).

Analysis

We provide descriptive statistics comparing the responses of different types of hospital staff. Statistically significant differences in responses by type of staff are identified using Fischer's Exact Tests. We then predict the probability of intervening among staff who witnessed parent-to-child hitting using logistic regression.

Results

Pooled Sample Description

Sample characteristics are shown in Table 1. Our respondents are 2,580 staff from the general medical center and 733 staff from the children's medical center. This constitutes about 39% of the general hospital staff and 10% of the children's hospital staff. The pattern of effects described below was generally the same for both medical centers included in the study. We tested for differences between medical centers for the descriptive analyses using Fischer's Exact Tests and found few differences (generally speaking, the children's medical center staff were more likely to report that they witnessed hitting and nurses and other direct care staff at the children's medical center were more likely to intervene than their counterparts at the general medical center). For the logistic regression models, we used Chow Tests to determine whether there were differences in coefficients across hospitals and found that there were not. Therefore, we pooled the samples for all results described below, though we continued to control for medical center in regression models. Exclusion for incomplete data (excluded n = 399 and n = 51 for the general medical center and children's hospital, respectively) resulted in a final sample of 2,863. Excluded observations were primarily respondents who did not report their type of job. The majority of the sample was female, White, and had some education beyond high school, although physicians and non-direct care staff were less likely to be female and White. Staff positions were categorized into four groups: (1) physicians; (2) nurses, including nurse practitioners; (3) other direct care staff (e.g., medical assistants, social workers); or (4) non-direct care staff (e.g., custodians, receptionists, medical records workers, lab technicians). Physicians, nurses and other direct care staff also tended to be somewhat younger than non-direct care staff; they were more likely to be in the 25-35 age range and less likely to be in the 46-55 age range. Physicians and nurses were more likely to be parents than other direct care staff and non-direct care staff.

Table 1. Sample Description (percentages).

	Physicians (n=115)	Nurses (n=878)	Other direct care staff (n=623)	Non-direct care staff (n=1247)
Female	61.4	94.9	87.3	78.8
White	84.4	95.8	92.0	90.4
Age 18-24	0.9	4.5	11.0	6.5
Age 25-35	34.2	30.4	34.0	22.9
Age 36-45	30.7	21.6	21.9	22.2
Age 46-55	14.4	23.1	19.2	28.3
Age 56 +	20.2	20.5	14.0	20.1
Parent	76.5	70.6	63.6	64.9

Witnessing Parent-to-Child Hitting

The frequencies of how often staff witness parents hitting children in medical center settings are presented in Table 2. Physicians were significantly more likely than the other three staff groups to report witnessing parent-to-child hitting at least a few times per year, with 1 in 2 physicians witnessing hitting compared with 1 in 4 nurses, 1 in 4 other direct care staff, and 1 in 6 non-direct care staff. Very few medical staff members (1 to 2%) reported witnessing parent-to-child hitting every month. Across all staff categories, 23% of the full sample, or 658 medical center staff, had witnessed parents hitting their children in the medical centers.

Table 2. Extent of medical center staff witnessing and intervening in cases of parent-to-child hitting in the medical center (in percentages).

	Physicians (n = 115)	Nurses (n = 878)	Other direct care staff (n = 623)	Non-direct care staff (n = 1247)
Percent of staff who witnessed parents hit their children in the medical center in the past year
A few times each year	47.8b,c,d [38.8-57.0]	22.6a,d [19.9-25.4]	25.2a,d [21.9-28.8]	16.3a,b,c [14.3-18.4]
Almost every month	2.6 [0.8-7.8]	2.1 [1.3-3.2]	1.6 [0.9-3.0]	1.1 [0.7-1.9]
At least once in the past year	50.4b,c,d [41.3-59.5]	24.5a,d [21.9-27.6]	27.0a,d [23.5-30.4]	17.3a,b,c [15.4-19.6]
Of staff who witnessed parents hitting children in past three months (n = 464), percent who:
Never took action	20.0d [10.7-34.4]	30.2d [23.6-37.8]	32.1d [23.8-41.6]	61.6a,b,c [53.6-69.0]
Sometimes took action	42.2d [28.6-57.1]	37.7d [30.5-45.4]	31.1 [23.0-40.6]	23.8a,b [17.7-31.3]
Always took action	37.8d [24.8-52.8]	32.1d [25.3-39.7]	36.8d [28.1-46.4]	14.6a,b,c [9.8-21.2]
Of staff who took action (intervened) sometimes or always (n = 279), percent who:
Were comfortable with intervening	79.4 [62.3-90.0]	75.2 [66.2-82.4]	83.1 [72.4-90.2]	72.4 [59.4-82.5]
Felt intervention was effective	82.4 [65.5-92.0]	85.3 [77.3-90.8]	88.7 [78.9-94.3]	82.8 [70.6-90.5]
Intervened by
Talking to the parent(s)	100.0b,c,d [N/A]	77.9a,d [69.2-84.6]	70.8a [59.2-80.2]	58.6a,b [45.5-70.6]
Talking to the children	19.4 [9.4-35.9]	28.3 [20.7-37.4]	36.1 [25.8-47.9]	20.7 [12.1-33.2]
Notifying a supervisor	11.1c [4.2-26.5]	21.2c [14.6-29.8]	43.1a,b,d [32.1-54.8]	25.9c [16.1-38.8]
Contacting a social worker	33.3 [19.8-50.3]	32.7 [24.7-42.0]	19.4 [11.8-30.3]	24.1 [14.7-36.9]
Filing a report with child protective services	8.3 [2.7-23.4]	4.4c [1.8-10.3]	15.3b,d [8.6-25.7]	3.5c [0.8-13.0[
Contacting security	8.3 [2.7-23.3]	9.7 [5.4-16.8]	2.8d [0.7-10.6]	17.2c [9.5-29.4]

Note: Significance of group differences is based on Fischer's Exact tests.

^asignificantly different from doctors at p<.05

^bsignificantly different from nurses at p<.05

^csignificantly different from other direct care staff at p<.05

^dsignificantly different from non-direct care staff at p<.05

Intervention Behavior

Among those who witnessed hitting in the previous three months (n = 464), less than one-third of staff members in all direct care categories (physicians, nurses, and others) reported that they never took action, compared with almost two thirds (61.6%) of non-direct care staff members. Significantly higher proportions of physicians, nurses, and other direct care staff members reported taking action every time they witnessed hitting (37.8%, 32.1%, and 36.8%, respectively) compared with non-direct care staff members, of whom only 14.6% reported taking action every time.

Of those staff members who sometimes or always intervened when they saw a parent hitting a child in the medical center (n = 279), strong majorities in all job categories felt at least somewhat comfortable doing so and felt they were at least somewhat effective, with no statistical difference among job categories on either dimension. All physicians who intervened reported that they spoke to the parent directly about their behavior, making them significantly more likely to do so than nurses, other direct care staff, and non-direct care staff. Contacting a supervisor was least common among physicians (11.1%) and most common among other direct care staff (43.1%). Similarly, other direct care staff members were twice as likely as physicians and more than three times as likely as nurses or non-direct care staff to contact child protective services. In contrast, staff members who did not work in direct care were significantly more likely to contact security than their direct care counterparts (17.2% vs. less than 10% for each direct care staff category; p< .01). Staff members from all job categories were equally likely to report having talked to the children themselves and to have contacted a social worker.

Factors Associated with Intervening

We next examined whether any attitudinal or demographic factors predicted whether or not a staff member intervened. The results of regression models examining the odds of intervening are shown in Table 3. A one standard deviation (SD) increase in positive attitudes toward intervening doubled the odds of always or sometimes intervening (versus never). Those with more comfortable strategies to intervene were also more likely to always do so: a one SD increase in having comfortable strategies was associated 2.8 times higher odds of always intervening. In addition, non-direct care staff members were 83% less likely to always or sometimes intervene than were physicians. Those who always intervened did not differ from those who sometimes or never intervened on perceived coworker attitudes toward spanking or demographic characteristics.

Table 3. Logistic regressions predicting intervention behavior among those who witnessed parent-to-child hitting.

	Always or sometimes intervened (vs. never) n = 436		Always intervened (vs. sometimes) n = 265
	OR	95% CI	OR	95% CI
Support for staff intervening in parent-to-child hitting	2.31**	[1.35 3.95]	1.49	[0.82 2.71]
Has comfortable strategies to intervene	2.80***	[2.11 3.70]	1.78**	[1.26 2.51]
Own positive attitudes about spanking	0.81	[0.50 1.30]	0.60	[0.36 1.01]
Perceived coworker positive attitudes about spanking	1.34	[0.90 1.98]	1.01	[0.69 1.48]
Job category (reference group: Physicians)
Nurses	0.53	[0.19 1.46]	1.27	[0.51 3.14]
Other direct care	0.41	[0.14 1.19]	1.33	[0.51 3.47]
Non-direct care	0.17***	[0.06 0.46]	1.09	[0.40 2.99]
Demographic characteristics
Female	1.28	[0.64 2.56]	1.59	[0.71 3.55]
Nonwhite	0.79	[0.34 1.81]	0.79	[0.30 2.11]
Age	0.92	[0.75 1.13]	0.86	[0.69 1.09]
Parent	0.93	[0.54 1.59]	0.78	[0.42 1.44]

Note: Models control for medical center site.

^*p< .05,

^**p< .01,

^***p< .001.

We also compared those who always intervened with those who sometimes intervened, excluding those who reported that they never intervened (see last column in Table 3). Only one characteristic distinguished those groups: those who intervened always had more comfortable strategies to intervene than did those who only intervened sometimes. Neither job type nor attitudes toward intervening distinguished those who always intervened from those who sometimes intervened.

Staff-reported Reasons for Not Intervening

Staff who witnessed parent-to-child hitting but did not always intervene (n = 334) were asked whether one of four reasons explained why they did not intervene and were invited to write-in their own reasons. The percent of staff who reported each reason is displayed in Table 4; there were no differences across job types in reasons for not intervening, thus the percentages in Table 4 are for the combined sample. Nearly half of the staff respondents reported that the most common reason they did not intervene was that they did not know what to say or do to stop the parent. One third of staff reported being concerned the parent would be more angry or abusive to the child if they intervened. Other reasons staff offered for not intervening were that they did not find the parent's behavior objectionable and that they did not think it was their place to intervene.

Table 4. Staff-reported reasons for not intervening (n = 334).

“I was not sure what to say or how to stop the parent from hitting their child”	48.2%
“I was worried that the parent might get angrier and become more abusive to the child”	34.1%
“I was concerned the parent might threaten or harm me in some way”	12.6%
“Did not want to embarrass the parent”	10.8%
Other reasons (open-ended)
Did not think parent was doing anything wrong	6.6%
Not my place to intervene	4.8%

Note: Respondents could choose more than one reason, thus percentages do not add to 100%.

Discussion

This study identified the prevalence of witnessed parent-to-child hitting in two medical centers as well as staff attitudes about and frequency of intervening when a parent hits a child. Physical punishment is an ineffective disciplinary technique associated with negative repercussions for child health and well-being and medical professionals, especially physicians, are trusted sources of information and advice on parenting behaviors. Thus, understanding the experiences and actions of medical personnel who witness parent-to-child hitting can provide critical insight into the potential role of the medical community in interventions targeting physical discipline. More than half of the physicians, about a quarter of the nurses and other direct care staff and 17% of non-direct care staff in this study witnessed at least one instance of a parent hitting a child in their medical center in the past year. Even if each of those staff only saw one incident, this would amount to 658 incidents of parent-to-child hitting per year across two medical centers—in other words, almost two incidents each day. These findings indicate that parent-to-child hitting is fairly common in medical center settings, and medical center staff may be in need of training in whether and how to intervene when it occurs.

There are many reasons for medical centers to want to reduce the incidence of parent-to-child hitting. Hitting children, including spanking, is recognized as harmful to children's physical, cognitive, and social-emotional development², and thus medical centers could play an important role in discouraging hitting children in the service of promoting children's health and well-being. It is also the case that the American Academy of Pediatrics and the National Association of Pediatric Nurse Practitioners have taken public positions against the use of physical punishment such as spanking.^5–7 The fact that parents tend to trust guidance from physicians on disciplinary practices¹¹ emphasizes the role that medical center staff can play in reducing parent-to-child hitting within medical settings and possibly beyond.

Despite the potential benefits of intervening, many staff who witnessed parent-to-child hitting on at least one occasion at their medical center never or rarely intervened. Failure to intervene was especially common among non-direct care staff (62%) and least common among physicians (20%). Along with job type, attitudes toward intervening and having comfortable strategies with which to intervene were the most consistent predictors of whether staff intervened each time they witnessed parent-to-child hitting. Educating staff about the empirical research on the adverse effects of hitting children may improve staff attitudes toward intervening. Some research has found that increasing knowledge about the ineffectiveness and harmfulness of physical punishment decreases approval of physical punishment^19,20, and attitudes toward physical punishment are strongly correlated with attitudes toward intervening in our sample.

Moreover, the reasons that staff did not intervene provide potential targets for future staff training efforts. Staff commonly reported that they did not intervene because they were not sure what to do (48%) and that they were afraid of parents' reactions (34%). Providing staff with clear strategies for how to intervene when they witness hitting is a relatively low-effort approach that could increase the rate of staff intervention. Ensuring that staff know how to engage parents in a supportive and nonjudgmental way may quell their concerns about negative parent reactions.

Educating staff about appropriate actions to take when parents hit their children would remove any ambiguity as to who is expected to take action and what actions are appropriate. One method currently being tried in a handful of medical centers throughout the country is the creation of a No Hit Zone, which makes clear to staff, patients, and visitors that hitting of all kinds is prohibited and which educates staff on how to intervene if they witness hitting.²¹ However, rigorous evaluation of these efforts has not yet occurred. Future research is encouraged to understand whether interventions, such as No Hit Zones, can bridge the gap between staff attitudes and actions and reduce the incidence of parent-to-child hitting in medical facilities.

It may also be important for medical centers to provide clear instruction to staff on when they should contact child protective services. As our study makes clear, staff rarely contact child protective services when they witness parent-to-child hitting. This may be because they only witnessed mild spanking, but it may be because staff are not sure what is permissible hitting and what constitutes physical abuse. Indeed, the legal distinction between physical punishment and physical abuse is a thin one; physical punishment that is “reasonable” is an exception to child abuse laws in 17 states, including Missouri.²² Thus, many professionals may be reluctant to adjudicate what is reasonable and instead err on the side of parental authority when they witness children being hit by parents. Yet, it is important to note that whereas mandated reporters who fail to report are at risk of legal consequence, any person who makes a report in good faith is immune from civil or criminal retribution.²³ In addition, hospitals may consider instituting reporting rules for non-direct care staff, who are not mandated to report child abuse in most states' laws¹⁴ but who, as this study demonstrates, do witness parent-to-child hitting while working in medical centers.

This study establishes that medical centers and hospitals are the settings for regular parent-to-child hitting and thus could also serve as key settings for interventions directed at parents. An example of such an approach is the Play Nicely program which provides education to parents in pediatric clinics and which has been found to be effective at reducing support for and intentions to use spanking.²⁴

We acknowledge that our study had some limitations. First, findings cannot be generalized to all medical centers and may be affected by selection bias. We only sampled two medical centers and, because the survey was entirely voluntary and there were no guaranteed financial or other incentives to participate, those who selected into the study (our respondents) may not be representative of their medical centers, nor of medical centers generally. Second, we had few physician respondents to the survey, which is significant given that physicians may be best able to influence parents' disciplinary practices. Third, staff may have difficulty recalling accurately whether they witnessed parent-to-child hitting and how they responded – a limitation of self-report data. Lastly, our sample had limited racial and gender diversity. Given that men and Blacks are somewhat more supportive of physical punishment than women or Whites²⁵, our study may have found more reluctance to intervene if it had included a more diverse sample.

Despite these limitations, this study is the first to document the presence of parent-to-child hitting in medical centers and to explore whether and in what ways staff respond. The prevalence of parent-to-child hitting documented here points to the need for training of medical center staff in how to respond and to an opportunity for educating both staff and parent clients about the potential harms of physical punishment.