Children's memory for pain is a robust predictor of future pain experience.31 Across clinical and experimental settings, healthy and illness populations, and early childhood and adolescence, it has been demonstrated that children who develop negatively biased memories of pain (ie, recalled pain is higher than initial report) report more pain and distress at future pain experiences.7,26,31 These memories are formed early in life and set the stage for how individuals cope with pain and approach/avoid medical care into adulthood.34 It has recently been shown that children's pain memories may underlie the development of chronic pain. Adolescents who developed more negatively biased recall of postsurgical pain reported higher levels of pain 4 months later, precisely at a time when pain can transition from an acute to a chronic state.31 Thus, understanding the processes underlying children's pain memory development is clinically important.
It has been said that “memory is not like a tape recorder”.49 Rather, memory is highly malleable and susceptible to distortion. This is particularly true in early childhood when children's memories are most susceptible to source errors and suggestibility effects.23 Having a researcher simply provide young children with brief language-based feedback after a needle procedure (ie, minimizing past pain and telling them that they were brave vs providing neutral feedback) can lead to fewer negative biases in pain memories.2 However, it is likely that how parents reminisce with their children about past painful events has the greatest influence on children's pain memory development. During early childhood, children begin engaging in extended conversations about the past with others,37 and parents play a powerful role in influencing how children understand, appraise, and re-appraise past autobiographical events, including events involving pain.29 Indeed, several decades of research in developmental psychology has shown that how parents reminisce with children about emotional events (eg, injuries and natural disasters) influences the accuracy of children's autobiographical memory development.45 Parents who are more elaborative (ie, provide new details) and who refer more to both positive and negative emotions while reminiscing have children who develop more accurate memories.44 Yet, no studies have investigated the influence of parent–child reminiscing about past painful events on the development of biases in children's recall of pain, which have been shown to underlie worse future pain outcomes.30 Although sex differences in adults' reminiscing have been found,41 research on differences in how mothers vs fathers reminisce with boys vs girls about past emotional events have yielded equivocal findings.3,12 While prior research has examined mothers' and fathers' verbal behavior during child pain22—finding no differences—research has not yet examined parent role and sex differences in parent–child reminiscing about past painful events and their influence on children's pain memory development.
This prospective study examined the influence of parent–child reminiscing about a recent past surgery in the subsequent development of children's recall of pain. We hypothesized that elements of reminiscing that have been shown to lead to more accurate memories of other emotionally-salient events (elaborative style and emotional content45) would be linked to more accurate/positively biased recall of pain. Given the dearth of literature on parent role and sex differences in reminiscing, we did not have specific hypotheses about how reminiscing would differ based on these individual differences.
This is the third manuscript from this longitudinal study. The first article examined the role of preoperative anxiety in children's memory biases.11 The second article examined differences in parent–child reminiscing for sad vs painful autobiographical events (ie, a tonsillectomy and another painful event).35 This second article was driven by a long tradition in the developmental psychology literature to examine differences in how parents reminisce with young children about different kinds of emotional events. It did not include a follow-up timepoint or assessment of children's recall of pain. Moreover, the aim of this second article was not to examine reminiscing about a surgery on recall of that surgery and sex and parent role differences in those processes and outcomes. Rather, it was driven by a developmental question about fundamental differences in reminiscing about a variety of autobiographical events. This article is the first to examine the predictive roles of parent–child narrative structure and content in the subsequent development of children's pain memory biases. It is also the first to examine differences in reminiscing and recall of pain based on parent role (mothers vs fathers) and child sex (boys vs girls). Methods pertaining to the unique aims of the current study are reported below.
2.1. Participants and setting
Between June 2016 and June 2018, research staff contacted and screened families whose children were scheduled to undergo a tonsillectomy with or without adenoidectomy through the ear, nose, and throat clinic at a tertiary children's hospital in Western Canada. Of the families contacted, 131 participants agreed to participate, 3 families did not complete baseline measures, 4 did not complete ≥ 2 after surgery questionnaires, 4 dropped out, and 8 were lost to follow-up (Fig. 1). No significant differences were revealed between families who completed vs dropped out from the study (ps > 0.05).
One-hundred twelve children (60% boys, Mage = 5.3 years, SD = 1.1) and one of their parents (34% fathers) participated in the study. Children were excluded from the current study if they received premedication with anxiolytics (eg, midazolam), had a developmental disability, a psychiatric diagnosis, or an American Society of Anesthesiology physical status ≥III. As noted above, 3 parents did not complete the initial questionnaire, 4 parents did not complete at least 2 postsurgical surveys; therefore, these dyads were excluded from the analyses. Data from 112 parent–child dyads, who completed all data points, were included in the analyses. Sociodemographic characteristics of the sample and family medical history are presented in Table 1. Independent-sample t-tests, χ2 tests, and omnibus tests (analyses of variances [ANOVAs]) did not reveal any significant differences on key variables based on sociodemographic variables. Ethical approval of the study was obtained from the institutional health research ethics board.
Parents were contacted by telephone approximately 1 to 3 weeks before surgery. If interested and eligible to participate, parents were sent a secure email link to the consent form and the baseline questionnaire through REDCap, a secure online survey web application.13 To participate, parents confirmed that they would be with their children during the postsurgical period and at each of the follow-up timepoints over the course of 1 month. Assent was obtained from children. On the day of surgery (DOS), a standard protocol for anesthesia was followed for each participant. An inhalational induction was employed using sevoflurane, oxygen, and/or nitrous oxide. After induction, anesthesia was maintained for the procedure with either total intravenous anesthetic or volatile anesthetics. All patients received analgesia of morphine, dexamethasone 0.2 mg/kg IV, and ondansetron 0.1 mg/kg IV during the procedure.
Consistent with previous research,19 on the DOS, parents received hard copies of the faces pain scales with verbal and written instructions on the back to assess their children's pain intensity and pain-related fear at home on the days after surgery. Parents received daily surveys through REDCap at 5:00 PM on days 1, 2, 3, 7, and 14 after surgery and indicated their children's report of pain intensity and pain-related fear for that particular day.
Two weeks after surgery, families were invited back to the hospital (where the research laboratory is physically located) for a laboratory visit. The visit involved a structured narrative elicitation task43 in which parent–child dyads were instructed to reminisce together about the recent tonsillectomy (ie, when children first came to the hospital for their surgery and the first few days after surgery). Consistent with previous research on parent–child reminiscing,42,43 parents and children were instructed by the researcher to talk as normally as possible, and for as long as they wished. During the parent–child narratives, the researcher left the room. Narratives were audio recorded and transcribed verbatim. After completion of the laboratory visit, parents and children received a $20 gift card to thank them for their time.
Approximately one month after surgery, a trained research assistant conducted an established memory interview26 with children through telephone using the same faces pain scales previously administered (FPS-R, Children's Fear Scale [CFS]). Specifically, children recalled their pain intensity and pain-related fear on the day of the surgery (DOS), day 1 after surgery, and days 1 to 3 after surgery (ie, an average pain rating for these 3 days). Although parents received instructions to assist their child in looking at the correct scale, they were specifically asked not to prompt their child, to facilitate assessment of what their child remembered independently. Each face on the scale corresponded to a randomly chosen letter underneath to avoid confounding a number with a face and to facilitate ease of communication through telephone.26 Upon completion of the study, participants received a $20 gift card to thank them for their time.
2.3.1. Parent measures
Parents reported their own and their child's age, sex, and ethnicity, medical history (ie, previous surgeries), and preparation for their child's tonsillectomy. Parents also reported their annual household income, education, and employment status.
2.3.2. Child measures
184.108.40.206. Pain intensity
The Faces Pain Scale Revised is a recommended self-report measure of children's pain intensity and was used to assess children's postsurgical pain intensity and memory for pain intensity (FPS-R).14 The FPS-R is a single-item measure depicting 6 gender-neutral faces ranging in facial expressions from neutral (0) to extreme pain (10). Children were asked by a researcher (at the hospital) or a parent (at home) to indicate how much they hurt by pointing to a face on the FPS-R. The FPS-R shows evidence of good test–retest reliability and construct validity for children in this age range47 and has previously been used to assess young children's recall of pain.27
220.127.116.11. Pain-related fear
The CFS was used to assess children's postsurgical pain-related fear and recall of postsurgical pain-related fear.21 The CFS comprises 5 faces displaying varying amounts of fear anchored by “not at all scared” (0) to “most scared possible” (4). The CFS has demonstrated good test–retest reliability and construct validity21 and has previously been used with young children to assess experienced and recalled pain-related fear.27 In line with a multifaceted approach to measuring children's pain memories,33 both the FPS-R and CFS were used to capture the sensory (pain intensity) and affective (pain-related fear) aspects of children's recall of pain.
2.3.3. Narrative coding scheme
Parent–child narratives from the laboratory visit were transcribed verbatim, broken into utterances (ie, sentences), and subsequently coded, based on adapted coding schemes drawn from the developmental psychology literature.43 Specifically, parent utterances were coded based on structure and content (for a full description of the adapted coding scheme with examples see Ref. 35). Consistent with previous research, proportions of each code type were calculated in relation to the total number of codes, to control for potential differences between dyads and narrative length.43 Two coders learned the coding scheme by concurrently coding 5 transcripts. After resolving discrepancies, one coder coded the remaining transcripts. Twenty percent of randomly chosen narratives (22) were coded by both coders to calculate reliability. Inter-rater reliability was calculated using Cohen's kappa and was excellent (ie, 0.85 for structure and 0.82 for content). Proportions of each narrative code type are presented in Table 3.
18.104.22.168. Narrative structure
Structural narrative coding was drawn from research by Sales et al.43 First, each parent utterance was coded as either a statement or a question. Then, questions were coded as being open-ended or close-ended (ie, yes–no questions). Both statements and questions could be further coded as repetitive (ie, containing old information from the conversation) or elaborative (ie, containing new information). One-word utterances were coded as evaluations (ie, “Yes” or “No”). Child utterances were coded based on whether they contained new information (ie, elaboration) or if they were a one-word reply (eg, “Yes,” “No,” “I don't know”). A child's utterance was coded as an off-topic switch if it contained unrelated content to the conversation. No codes were provided for utterances after an off-topic switch.
22.214.171.124. Narrative content
Research by Sales and Fivush informed the content coding scheme used in the current study.43 Parent and child utterances were coded based on whether they contained words that fell into the following content categories: pain, coping, explanation, anxiety/fear, and emotion (negative, neutral, and positive). As such, utterances could receive multiple codes (eg, if being scared was mentioned in an utterance, it would receive 2 codes: “anxiety/fear” and “negative emotion”). Proportions of each type of content code were calculated in relation to the total amount of content codes.
2.4. Data analyses
Statistical analyses were performed using SPSS v. 24.15 Descriptive statistics were conducted to characterize the sample and obtain mean values and SDs of the key variables. Independent-sample t-tests, χ2 tests, and omnibus tests (ANOVAs) were used to examine SES differences on the key variables, as well as differences between families who completed vs withdrew from the study.
2.4.1. Narrative codes and biased recall of pain
Similar to previous research,26,32 children's initial ratings of pain and pain-related fear (DOS, day 1 after surgery, and an average of day 1-3 ratings) were used as covariates in analyses to assess biases in children's recalled pain. Specifically, the initial pain rating that corresponded to the recalled pain rating was entered as a covariate to determine the degree to which recall was biased (eg, the DOS pain intensity rating was entered as a covariate in analyses pertaining to recalled DOS pain intensity). If pain intensity or pain-related fear ratings for 1 of the 3 days (5% of pain intensity and 4% of pain-related fear) were missing, an average of 2 ratings was used. A series of partial correlations with 2-tail hypothesis testing was conducted to first examine the associations between surgery narrative codes and children's recall biases for pain intensity and pain-related fear. Narrative codes that were significantly associated with biases in children's recall of pain were used as predictors in the hierarchical regression analyses. Child age, sex, and initial ratings of pain intensity and pain-related fear were entered in the first 2 steps of every model.
2.4.2. Sex and parent role differences
First, to examine differences based on child sex and parent role, multivariate ANOVAs (MANOVAs) were conducted to examine whether parents and children differed in their use of narrative codes. Significant MANOVAs were followed up with independent-sample t-tests with Bonferroni-adjusted alphas. To examine whether the interaction between child sex and parent role influenced parent and child use of narrative codes, factorial ANOVAs were used. Significant factorial ANOVAs were followed up with simple effect tests (ANOVAs), and Bonferroni alpha adjustment was applied.
Finally, analyses of covariance were conducted to assess sex and parent role differences in children's memory biases for pain and pain-related fear with initial pain ratings included as covariates.
3.1. Descriptive statistics
Descriptive statistics of the key variables are presented in Table 2.
3.1.1. Medical history and hospital stay
Most children (67%) had not received a previous surgery, whereas 72% of parents had undergone a surgical procedure (40% had a tonsillectomy). Sixty-eight percent of parents sought additional information about the surgery (eg, searched for information online, attended a Surgery 101 workshop). Most children (62%) presented with an obstructive sleep apnea diagnosis. Nearly all participants (98%) received an inhalational anesthesia with sevoflurane (95%). On average, children received 2.5 mg (SD = 1.4) of intraoperative morphine. Most children (90%) were extubated deep. All children were admitted to the Pediatric Acute Care Unit and stayed for an average of 47.1 minutes (SD = 21.3). While in Pediatric Acute Care Unit, 32% of children received morphine (M = 1.3 mg, SD = 0.9) and 19% received Tylenol (M = 387.2 mg, SD = 383.5). On average, children reported 2.9/10 (SD = 3.4) pain intensity and 0.8/4 pain-related fear (SD = 1.3) ratings 2 to 3 hours after surgery (DOS ratings). Most children went home the same day of their surgery (72%) and did not have any postsurgery complications (77%). There were no significant differences revealed on any of the key variables (ps > 0.05) between children who went home vs who did not go home on the DOS. Moreover, no differences were found between children who did vs who did not experience any postsurgical complications.
3.1.2. Postsurgical and recalled pain and pain-related fear ratings
On day 1 after surgery, children reported 3.9/10 (SD = 3.2) pain intensity and 0.8/4 (SD = 1.3) pain-related fear ratings. On days 1 to 3, children reported 3.7/10 (SD = 2.4) pain intensity and 0.8/4 (SD = 1.0) pain-related fear ratings. One month after the tonsillectomy, children recalled 4.6/10 (SD = 3.6) pain intensity and 1.4/4 (SD = 1.4) pain-related fear immediately after the surgery, 4.6/10 (SD = 3.7) pain intensity and 1.4/4 (SD = 1.4) pain-related fear on day 1 after the surgery, and 4.5/10 (SD = 3.7) pain intensity and 1.3/4 (SD = 1.4) pain-related fear on days 1 to 3 after the surgery.
3.1.3. Narrative codes used during the laboratory visit and the memory interview
On average, participants completed the narrative elicitation task 16.6 days after the surgery (SD = 4.3, range = 9-35 days). During that interim period (ie, from after the surgery to the time of the narrative elicitation task), parents reported talking about the tonsillectomy a moderate amount (M = 4.9/10; SD = 2.6, 0 = “not at all,” 10 = “a lot”). On average, parents produced 78.6 utterances (SD = 50.2), and children responded with 45.7 utterances (SD = 30.7). The memory interview took place 32 days after the surgery (SD = 5.3, range = 25-50 days). Mean values and SDs of narrative codes are presented in Table 3.
3.2. Recall biases and narrative codes: correlational analyses
A series of partial bivariate correlations controlling for initial pain ratings revealed significant associations between narrative codes used by parents and children's recall of pain. Specifically, a higher proportion of statement elaborations (SEs) (ie, utterances containing novel information: SE) was associated with more positively biased recall of days 1 to 3 pain-related fear (r = −0.21, P = 0.023). Higher levels of overall parent use of elaboration elements (ie, narrative structure code containing novel information: MQE, YNE, and SE) were associated with more positively biased recall of pain-related fear on day 1 after surgery (r = −0.20, P = 0.038). Conversely, a higher proportion of parent evaluations were associated with more negatively biased recall of DOS pain intensity (r = 0.20, P = 0.036) and recalled pain-related fear on day 1 after surgery (r = 0.26, P = 0.006). Other elements of parent reminiscing style and the child narrative codes were not significantly associated with biases in children's recall of pain (ps > 0.05).
With regard to parent content narrative codes, a higher proportion of words related to positive emotions and emotions in general (ie, negative, positive, or neutral emotions) was correlated with more positively biased recall of days 1 to 3 pain intensity (r = −0.21, P = 0.025; r = −0.23, P = 0.014), respectively. A higher proportion of pain-related words was associated with more negatively biased recall of DOS pain intensity (r = 0.32, P = 0.001) and pain-related fear (r = 0.22, P = 0.018).
With regard to child content narrative codes, a higher proportion of overall emotion-laden words (ie, words related to negative, positive, or neutral emotions) was associated with more positively biased recall of pain intensity on day 1 after surgery (r = −0.25, P = 0.008). More frequent use of explanation-related words was also associated with more positively biased recall of DOS pain intensity (r = −0.21, P = 0.028). Similar to parent use of pain-related words, a higher proportion of pain-related words was correlated with negatively biased recall of DOS pain intensity (r = 0.28, P = 0.003). Other narrative content codes were not significantly associated with biases in children's recall of pain (ps > 0.05).
3.3. Recall biases and narrative codes: hierarchical regression analyses
In line with previous research,32 the hierarchical models included predictors (narrative codes) and outcomes (recalled pain intensity or recalled pain-related fear) that were significantly correlated. Four hierarchical regression models were conducted to examine the influence of narrative structure (ie, parent elaboration style) as well as parent and child narrative content on biases in children's recall of pain intensity and pain-related fear. A fifth model examined the influence of multiple narrative codes on biases in children's recall of DOS pain. Child characteristics (age and sex) and initial pain ratings that corresponded to the recall ratings were controlled in the first and second steps of the models, respectively. The results of these analyses are summarized in Table 4.
3.3.1. Model 1
126.96.36.199. Parent narrative elaboration style and recall of pain-related fear (days 1-3)
Child age, sex, and initial report of pain-related fear (days 1-3) accounted for 9.6% of the variance in children's recall of pain-related fear (days 1-3) (F(3, 111) = 3.93, P = 0.01), with child age and sex accounting for 1.5% (ΔF(2, 112) = 0.87, P > 0.05) and the initial pain report accounting for 8.1% of the variance (ΔF(1, 111) = 9.92, P < 0.01). Above and beyond child age, sex, and initial reports of pain-related fear, the proportion of parent SE) significantly accounted for 3.3% of the variance in children's recall of pain-related fear (days 1-3) (ΔF(1, 110) = 4.20, P < 0.05). A negative beta weight suggests that greater use of parent SEs was related to more positively biased memories of pain-related fear (days 1-3).
3.3.2. Model 2
188.8.131.52. Parent narrative elaboration style and recall of pain-related fear (day 1)
Child age, sex, and initial report of pain-related fear (day 1) accounted for 7.4% of the variance in children's recall of pain-related fear (day 1) (F(3, 110) = 2.94, P < 0.05), with age and sex accounting for 4.8% (ΔF(2, 111) = 2.82, P > 0.05) and the initial report accounting for 2.6% of the variance (ΔF(1, 110) = 3.08, P > 0.05). The proportion of parent overall elaborations (MQE, YNE, and SE) accounted an additional 2.2% of the variance in children's recall of pain-related fear (day 1) (ΔF(1, 109) = 2.65, P > 0.05), while controlling for child sex, age, and initial reports of pain-related fear (day 1). A negative beta weight suggests that greater parental use of overall elaborations was related to more positively biased recall of pain-related fear (day 1).
3.3.3. Model 3
184.108.40.206. Parent narrative content and recall of pain intensity (days 1-3)
Child age, sex, and initial report of pain intensity (days 1-3) accounted for 6% of the variance in children's recall of pain intensity (days 1-3) (F(3, 109) = 2.34, P > 0.05), with age and sex accounting for 0.8% (ΔF(2, 110) = 0.42, P > 0.05) and the initial report accounting for 5.3% of the variance (ΔF(1, 109) = 6.14, P < 0.05). Above and beyond child characteristics and initial reports of pain intensity, the proportion of parent-used words associated with positive emotions significantly accounted for an additional 4.6% of the variance in children's recall of pain intensity (days 1-3) (ΔF(1, 108) = 5.53, P < 0.05). A negative beta weight suggests that greater parental use of positive emotional language was related to more positively biased recall of pain intensity (days 1-3).
3.3.4. Model 4
220.127.116.11. Child narrative content and recall of pain intensity (day 1)
Child age, sex, and initial report of pain intensity (day 1) accounted for 16.1% of the variance in children's recall of pain intensity (day 1) (F(3, 107) = 6.85, P < 0.001), with age and sex accounting for 2.6% (ΔF(2, 108) = 1.45, P > 0.05) and the initial report accounting for 13.5% of the variance (ΔF(1, 107) = 17.23, P < 0.001). The proportion of emotion-laden words used by children additionally accounted for a significant proportion (5.6%) of the variance in children's recall of pain intensity (day 1) (ΔF(1, 106) = 7.54, P < 0.01). A negative beta weight suggests that greater child use of emotion words was related to more positively biased recall of pain intensity (day 1).
3.3.5. Model 5
18.104.22.168. Parent and child narrative content and recall of pain intensity
Multiple narrative codes (ie, parent use of evaluations and pain-related words, children's use of pain- and explanation-related words) were significantly related to biases in children's recall of DOS pain and were included as predictors in Model 5. Child age, sex, and initial report of DOS pain accounted for 1.4% of the variance in children's recall of DOS pain (F(3, 108) = 0.51, P > 0.05), with age and sex accounting for 1.1% (ΔF(2, 109) = 0.62, P > 0.05) and the initial report of DOS pain intensity accounting for 0.3% of the variance (ΔF(1, 108) = 0.30, P < 0.05). Above and beyond child age, sex, and initial reports of pain intensity, narrative codes significantly accounted for 17.1% of the variance in children's recall of DOS pain intensity (ΔF(4, 104) = 5.47, P < 0.001) with the proportion of parent-used words associated with pain being the only significant narrative code predictor. A positive beta weight suggests that greater parental use of pain words was related to more negatively biased recall of pain intensity (DOS).
22.214.171.124. Parent role and sex differences in narrative codes and recall biases
Multivariate ANOVAs and analyses of covariance did not reveal any significant differences in recall biases or narrative styles between boys and girls (ps > 0.05). Similarly, children of fathers vs mothers did not differ in their recall of pain (ps > 0.05). However, parents significantly differed in their use of narrative codes. Specifically, fathers (M = 0.09, SD = 0.07) used explanations more frequently compared with mothers (M = 0.06, SD = 0.06, t(113) = 2.33, P = 0.021, η2 = 0.044). Children of fathers (M = 0.14, SD = 0.14) vs mothers (M = 0.09, SD = 0.10) used more words related to negative emotions (t(111) = 2.51, P = 0.014, η2 = 0.030).
Furthermore, we performed between-subject 2 (child sex: boys and girls) × 2 (parent role: mother and father) ANOVAs to investigate the effects of child sex and parent role on parent–children use of narrative codes. Parent use of negative emotion words was qualified by a significant interaction between child sex and parent role, F(1, 114) = 8.18, P < 0.01, η2 = 0.069. Follow-up t-tests revealed that fathers used negative emotion words similarly with boys and girls (P > 0.05), whereas mothers used negative emotion words more frequently with boys (M = 0.13, SD = 0.09) than girls (M = 0.09, SD = 0.06), t(73.52) = 2.71, P < 0.01. Children's use of pain-related words was also qualified by a significant interaction between child sex and parent role, F(1, 113) = 6.73, P < 0.05, η2 = 0.058. Boys used pain-related words with the same frequency when talking with fathers and mothers (P < 0.05). Girls, however, used more pain-related words when reminiscing with mothers (M = 0.15, SD = 0.15) as compared to fathers (M = 0.06, SD = 0.07), t(44) = 2.57, P < 0.05.
This study examined the roles of mother– and father–child reminiscing about a past surgery on young children's subsequent pain memory development. Findings revealed that how parents and children talked about this past painful event was related to the development of subsequent biases in children's recall of pain. Parents who used a more elaborative reminiscing style had children who developed more accurate/positively biased recall of pain-related fear. Greater use of emotion-laden words by parents (ie, positive) and children (ie, negative, positive, and neutral) while reminiscing about the past surgery was related to children developing more accurate/positively biased recall of pain intensity. Conversely, when parents used more pain-related words while reminiscing about the past surgery, children tended to develop more negatively biased recall of pain. Taken together, these findings underscore the importance of parent-child reminiscing about past painful events in influencing children's subsequent pain memory development and begin to isolate specific narrative elements that are linked to negative biases in children's pain memories.
These findings are in keeping with both the pediatric pain and developmental psychology literature. Those elements of parent–child reminiscing about a past surgery that were linked to more accurate and positively biased memories for pain have also been shown to be linked to other adaptive developmental outcomes. Indeed, a more elaborative style of reminiscing about past negative emotional events has been shown to be robustly linked to better cognitive (eg, memory and language), social (eg, prosociality), and emotional (eg, emotion regulation) developmental outcomes.17,18,45 Elaborative reminiscing style creates coherent, story-like narratives and elucidates causal relationships associated with the past event.36 Reminiscing about past emotions, including negative emotions, enhances children's emotion knowledge and understanding, which, in turn, may lead to better self- and emotion-regulation skills.16,18 The current findings extend this literature by demonstrating that parent–child reminiscing about a past painful experience is also linked to the development of more adaptive (ie, more accurate or positively biased) pain memories. The finding that greater parental use of pain words while reminiscing about the past surgery was linked to more negatively biased recall of pain is also in line with previous research on acute pediatric pain. Several empirical studies have shown that parents who attend more to pain while talking to children immediately before and during acutely painful experiences (needle procedures and experimentally induced pain) have children who experience greater immediate pain and distress.4,5,20,50 This is likely due to the fact that drawing children's attention to pain heightens their somatic sensations, threat perception, and bodily vigilance.50 Our findings extend this research by showing that a linguistic emphasis on the sensory and affective aspects pain while reminiscing about a past painful experience has a long-lasting influence on the development of children's memories for pain.
Although boys and girls did not differ in terms of recall biases, reminiscing differed between mother– and father–child dyads. Fathers used more explanations when reminiscing with their children than mothers. In addition, children of fathers, but not mothers, used more words related to negative emotions. Mothers, but not fathers, used more negative emotion words when talking with boys vs girls. Moreover, girls, but not boys, used more pain-related words when reminiscing with mothers vs fathers. Although these results are preliminary and the narrative codes that did differ were not related to recall biases for pain, children of parents who use more explanations and refer more to emotions while reminiscing have been shown to have better developmental outcomes (eg, emotional understanding and accurate memories).45 Research should examine other future pain, health, and developmental outcomes to gain a greater understanding of the degree to which these narrative differences are clinically significant and may be a pathway for the socialization of pain behaviors. It has been argued that the potential novelty of talking to fathers about past autobiographical events and greater attempts to be engaging with them may translate into children using more emotion-laden words with their fathers vs mothers.1,40 Moreover, when talking to young children, fathers have been shown to focus relatively more on what happened vs the emotional content of the event,12 which could account for why fathers explained more to children while reminiscing than mothers did. Further investigations of parent role and sex differences in reminiscing using more advanced statistical analyses (eg, dyadic, sequential, and contingency responding analyses) and additional longitudinal outcomes are warranted. In particular, future research should use time-window sequential analysis8 to examine the bidirectional relationships between parent and child utterances as well as the interactions between them in influencing outcomes (including memory and pain).
From a clinical perspective, we do not believe that this research suggests that parents should not reminisce with their children about pain. Rather, it points to how parents may most adaptively reminisce about past painful experiences to potentially buffer against children developing negatively biased pain memories. By using an elaborative reminiscing style, parents engage their children in a rich discussion about their past experience, filling in new details, encouraging and fostering a coherent narrative about this past experience, and also co-constructing the meaning of that experience. Moreover, talking about painful experiences need not overfocus on the sensory and affective aspects of pain itself but rather emphasize other aspects of the overall experience. The current findings suggest that how parents reminisce (eg, using an elaborative reminiscing style) and what parents talk about (eg, using words related to positive emotions and pain) with their children is what is predictive of biases in children's pain memories. Given the malleability of memory, particularly during early childhood,29 and the pivotal role of parental reminiscing on children's cognitive development during this period of development,24 this is likely an ideal time to intervene.25 Even among high-risk samples of parents (eg, low SES), interventions to teach parents how to more elaboratively reminisce with their children about past autobiographical events have been efficacious in fostering better developmental (including cognitive) outcomes.45 Moreover, parents have been effectively trained to use less pain-attending language with their children during acute experimental pain to improve immediate pain outcomes.5,51 Thus, it is likely that parents could be effectively taught how to reminisce with their children in ways informed by this study (eg, more elaborative style, less-pain words, and more references to positive emotions) to foster more adaptive pain memory development; however, such a trial has not yet been conducted. Furthermore, while the current study focused on parents' reminiscing style, we think that this work also has relevance to other adults in children's lives, including their health care providers. Indeed, past research that has examined memory reframing interventions involving talking to children in a particular way about their past painful procedures has used adult researchers reminiscing with children about painful experiences before imminent needle procedures.28 Health care professionals could reminisce with children about past procedural experiences before procedures in ways that are elaborative, de-emphasizing details specific to pain sensations, and pulling for information that involved positive emotions (eg, a friendly nurse, helpful strategies used, getting a treat afterward, etc.). This is an important area for future research.
The clinical significance of the findings is supported by previous research linking similar/any deviations in children's memories for pain to ratings of distress and pain during future pain experiences. This has been demonstrated across populations (cancer and healthy) and pain contexts (needles and surgeries), which support the clinical significance and robustness of these findings.7,26,32 The current study did not examine outcomes beyond assessment of children's recall; therefore, we draw information about the clinical significance of the current findings and operational definitions from past research relating memory biases of similar magnitude to pain, fear, and distress at subsequent pain experiences. Among healthy 8- to 12-year-old children undergoing cold pressor pain and adolescents undergoing surgeries such as pectus repair and spinal fusion, the effect sizes for the relationships between memory biases and subsequent pain and fear were moderate to high (ie, correlation coefficients within the range of 0.2-0.626,32). In more complex regression models that controlled for child age and sex, initial pain ratings, anxiety, and children's memories for cold pressor pain explained 15% of variance in subsequent pain report.26 Finally, a medium effect size (b = 0.419; ß = 0.415) has been found for the relationship between children's memories for postsurgical pain on future pain 4 months after surgery.31 We believe that the findings of the current study lay the foundation for the development of clear and easily developed interventions to target parent–child reminiscing to foster more accurate and positively estimated memories for pain. Based on previous trials of memory reframing interventions for needle pain6,38 that have found moderate effect sizes for the impact of the intervention on recall, we would hypothesize similar effect sizes. However, to empirically demonstrate this, it is necessary for future research to build upon this detailed observational work and conduct a controlled trial to determine the impact of parent–child reminiscing on children's recall of pain.
This study had limitations and underscores several areas for future research. First, this study did not capture nonverbal information or the affective tone used during parent–child reminiscing about past pain. Research suggests that the positive or negative influence/interpretation of a specific use of language (eg, reassurance) in the context of acutely painful procedures may be dependent on parental tone of voice and facial expression.20 Future research should apply a broader assessment of verbal and nonverbal information in the narratives. In addition, we acknowledge the uniqueness of this type of surgery and the potential lack of generalizability to other types of acutely painful procedures. Tonsillectomies occur at only one point in a child's life and therefore are both novel and salient, 2 factors that influence memory consolidation and recall.46 While from a methodological perspective, this is optimal because recall of pain can more reliably be tied to a specific time and place, it is unknown how parent–child reminiscing and its influence on children's memories would differ for repeated painful events (eg, vaccine injections). Indeed, the accuracy of children's recall has been shown to differ for single vs repeated autobiographical events.39 Furthermore, parent–child dyads reminisced about the entire postsurgical experience, which lasted over the span of several days and included many aspects (eg, relatives and friends visiting, not going to school, and disruptions to their usual routine) in addition to the experience of pain. Repeated painful events (eg, vaccine injections) are shorter, occur every year or more frequently, and are more centered around the pain and fear experience. Furthermore, despite active advocacy efforts, pain management during needle procedures is often inadequate,48 whereas after surgery (ie, during the in-hospital stay and the first 24 hours after discharge), most children receive pharmacological and/or nonpharmacological pain management.10 Examining the influence of parent–child reminiscing about repeated painful events and its role in the development of children's pain memories and future pain experience is a key avenue for future research. Research is needed to understand how the tonsillectomy experience generalizes to other clinical pain contexts.
Although the current findings show that an elaborative reminiscing style was linked to adaptive (ie, less negatively biased) memories, it could be argued that elaboration, and specifically, the tendency to introduce new content to the conversation could be construed as a tendency to avoid further discussion about a particular topic. We have demonstrated that parents tend to use elaboration more frequently when talking about past sad events vs painful events and have argued that one potential reason for this could be parental tendencies to avoid an uncomfortable, or potentially triggering topic of past pain.35 Previous work has shown that in analyzing parent–child conversations that mentioning pain, an overwhelming majority were about present pain or future/imaginary pain. Only 2% of parent–child dyads talked about past pain experiences.9 Further work is needed to test this hypothesis, and we believe this line of inquiry could be enriched by incorporating sociological and cultural perspectives. In addition, dispositional characteristics may influence parent–child reminiscing about painful events. In previous research with adolescents undergoing major surgeries, higher parent catastrophizing about child pain was found to be related to more negatively biased postoperative pain memories,32 and it is possible that parent–child interactions underlie this relationship. There may be other factors that may influence parent–child reminiscing about pain and recall biases, particularly in younger children. For example, the nature of the parent–child attachment relationship, the parents' own memories and experiences with similar types of pain and, more generally, parenting stress may all play a role. This is a key area for future research. Moreover, parents' memories of their own past surgeries were not assessed and could have influenced parent–child reminiscing and children's recall. In addition, the current sample was predominantly white, which precluded examination of cultural differences in reminiscing; however, this is an important avenue for future research. Finally, developmental studies1,40 have strived to include both parents of each child as reminiscing partners. Because of feasibility issues, we did not examine within-family differences in reminiscing; therefore, influences of either parent on their child's memory biases cannot be determined.
In conclusion, this study examined the influences of parent–child reminiscing about a past surgery on children's subsequent pain memory development. A more elaborative parental reminiscing style and more references to positive emotions in the narratives were linked to more accurate and positively biased recall of pain. Conversely, greater parental use of pain-related words was related to more negatively biased recall of pain. Although parent role and sex differences in children's pain memory biases were not found, several differences in reminiscing emerged. Fathers reminisced in ways (more negative emotions and more explanations) that have been previously linked to more adaptive developmental outcomes. Moreover, mothers used more negative emotional language with boys, and girls used more pain-related words when reminiscing with mothers. By isolating specific elements of parent–child reminiscing that are linked to biases in children's recall of pain, which are robust predictors of future pain experience,7,26,31 this work can inform the development of parent-led interventions to reframe children's memories for pain to be more accurate and positive.
Conflict of interest statement
The authors have no conflict of interest to declare.
This work was supported by the Society of Pediatric Psychology Targeted Research Grant, American Pain Society Future Leader in Pain Grant, Maternal Newborn Child and Youth Strategic Clinical Network Health Outcomes Improvement Fund, University of Calgary Faculty of Science Seed Grant as well as start up funding from the Alberta Children's Hospital Research Institute awarded to Noel. Pavlova was supported by the Alberta Strategy for Patient-Oriented Research Graduate Studentship. Graham's contributions were supported by funding through the Alberta Children's Hospital Foundation.
. Adams S, Kuebli J, Boyle PA, Fivush R. Gender differences in parent-child conversations about past emotions—a longitudinal investigation. Sex Roles 1995;33:309–23.
. Bruck M, Ceci SJ, Francoeur E, Barr R. “I hardly cried when I got my shot!” Influencing children
's reports about a visit to their pediatrician. Child Dev 1995;66:193–208.
. Buckner JP, Fivush R. Gendered themes in family reminiscing
. Caes L, Vervoort T, Devos P, Verlooy J, Benoit Y, Goubert L. Parental distress and catastrophic thoughts about child pain: implications for parental protective behavior in the context of child leukemia-related medical procedures. Clin J Pain 2014;30:787–99.
. Chambers CT, Craig KD, Bennett SM. The impact of maternal behavior on children
's pain experiences: an experimental analysis. J Pediatr Psychol 2002;27:293–301.
. Chen E, Zeltzer LK, Craske MG, Katz ER. Alteration of memory
in the reduction of children
's distress during repeated aversive medical procedures. J Consult Clin Psychol 1999;67:481–90.
. Chen E, Zeltzer LK, Craske MG, Katz ER. Children
's memories for painful cancer treatment procedures: implications for distress. Child Dev 2000;71:933–47.
. Chorney JM, Garcia AM, Berlin KS, Bakeman R, Kain ZN. Time-window sequential analysis: an introduction for pediatric psychologists. J Pediatr Psychol 2010;35:1061–70.
. Craig KD, Stanford EA, Fairbairn NS, Chambers CT. Emergent pain language communication competence in infants and children
. Enfance 2006;58:52–71.
. Finley GA, McGrath PJ, Forward SP, McNeill G, Fitzgerald P. Parents
' management of children
's pain following “minor” surgery
. PAIN 1996;64:83–7.
. Fischer S, Vinall J, Pavlova M, Graham S, Jordan A, Chorney J, Rasic N, Brookes JT, Hoy M, Yunker WK. The role of anxiety in young children
's pain memory
development following surgery
. PAIN 2018 In Press.
. Fivush R, Brotman MA, Buckner JP, Goodman SH. Gender differences in parent-child emotion narratives. Sex Roles 2000;42:233–53.
. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009;42:377–81.
. Hicks CL, von Baeyer CL, Spafford PA, van Korlaar I, Goodenough B. The Faces Pain Scale-revised: toward a common metric in pediatric pain
measurement. PAIN 2001;93:173–83.
. IBM Corp. IBM SPSS statistics for windows. Armonk: IBM Corp, 2016.
. Laible D. Mother-child discourse in two contexts: links with child temperament, attachment security, and socioemotional competence. Dev Psychol 2004;40:979–92.
. Leyva D, Berrocal M, Nolivos V. Spanish-speaking parent-child emotional narratives and children
's social skills. J Cogn Dev 2014;15:22–42.
. Leyva D, Nolivos V. Chilean family reminiscing
about emotions and its relation to children
's self-regulation skills. Early Educ Dev 2015;26:770–91.
. MacLaren JE, Kain ZN. Prevalence and predictors of significant sleep disturbances in children
undergoing ambulatory tonsillectomy and adenoidectomy. J Pediatr Psychol 2008;33:248–57.
. McMurtry CM, Chambers CT, McGrath PJ, Asp E. When “don't worry” communicates fear: children
's perceptions of parental reassurance and distraction during a painful medical procedure. PAIN 2010;150:52–8.
. McMurtry CM, Noel M, Chambers CT, McGrath PJ. Children
's fear during procedural pain: preliminary investigation of the Children
's Fear Scale. Health Psychol 2011;30:780–8.
. Moon EC, Chambers CT, McGrath PJ. “He says, she says”: a comparison of fathers' and mothers' verbal behavior during child cold pressor pain. J Pain 2011;12:1174–81.
. Nelson K. Development of extended memory
. J Physiol Paris 2007;101:223–9.
. Nelson K, Fivush R. The emergence of autobiographical memory
: a social cultural developmental theory. Psychol Rev 2004;111:486–511.
. Noel M. Harnessing the fragility of pain memories to help children
forget: a new avenue for pediatric psychology interventions? J Pediatr Psychol 2016;41:232–4.
. Noel M, Chambers CT, McGrath PJ, Klein RM, Stewart SH. The influence of children
's pain memories on subsequent pain experience. PAIN 2012;153:1563–72.
. Noel M, McMurtry CM, Chambers CT, McGrath PJ. Children
for painful procedures: the relationship of pain intensity, anxiety, and adult behaviors to subsequent recall. J Pediatr Psychol 2010;35:626–36.
. Noel M, McMurtry CM, Pavlova M, Taddio A. Brief clinical report: a systematic review and meta-analysis of pain memory
reframing interventions for children
's needle procedures. PAIN Pract 2018;18:123–9.
. Noel M, Palermo TM, Chambers CT, Taddio A, Hermann C. Remembering the pain of childhood: applying a developmental perspective to the study of pain memories. PAIN 2015;156:31–4.
. Noel M, Pavlova M, McCallum L, Vinall J. Remembering the hurt of childhood: a psychological review and call for future research. Can Psychol 2017;58:58–68.
. Noel M, Rabbitts JA, Fales J, Chorney J, Palermo TM. The influence of pain memories on children
's and adolescents' post-surgical pain experience: a longitudinal dyadic analysis. Health Psychol 2017;36:987–95.
. Noel M, Rabbitts JA, Tai GG, Palermo TM. Remembering pain after surgery
: a longitudinal examination of the role of pain catastrophizing in children
's and parents
' recall. PAIN 2015;156:800–8.
. Ornstein PA, Manning EL, Pelphrey KA. Children
for pain. J Dev Behav Pediatr 1999;20:262–77.
. Pate JT, Blount RL, Cohen LL, Smith AJ. Childhood medical experience and temperament as predictors of adult functioning in medical situations. Child Health Care 1996;25:281–98.
. Pavlova M, Graham S, Jordan A, Chorney J, Vinall J, Rasic N, Brookes J, Hoy M, Yunker W, Noel M. Socialization of pain memories: Parent-child reminiscing
about past painful and sad events. J Pediatr Psychol 2019. doi: [Epub ahead of print].
. Peterson C, McCabe A. Parental styles of narrative
elicitation: effect on children
structure and content. First Lang 1992;12:299–321.
. Peterson C, Ross A, Tucker VC. Hospital emergency rooms and children
's health care attitudes. J Pediatr Psychol 2002;27:281–91.
. Pickrell JE, Heima M, Weinstein P, Coolidge T, Coldwell SE, Skaret E, Castillo J, Milgrom P. Using memory
restructuring strategy to enhance dental behaviour. Int J Paediatr Dent 2007;17:439–48.
. Powell MB, Thomson DM. Children
of an occurrence of a repeated event: effects of age, repetition, and retention interval across three question types. Child Dev 1996;67:1988–2004.
. Reese E, Haden CA, Fivush R. Mothers, fathers, daughters, sons: gender differences in autobiographical reminiscing
. Res Lang Soc Interac 1996;29:27–56.
. Ross M, Holmberg D. Recounting the past: gender differences in the recall of events in the history of a close relationship. In: Zanna MP, Olson JM, editors. Self inference processes. Hillsdale: Lawrence Erlbaum Associates Inc, 1990.
. Sales JM, Fivush R. Social and emotional functions of mother-child reminiscing
about stressful events. Soc Cogn 2005;23:70–90.
. Sales JM, Fivush R, Peterson C. Parental reminiscing
about positive and negative events. J Cogn Dev 2003;4:185–209.
. Salmon K, Reese E. Talking (or not talking) about the past: the influence of parent-child conversation about negative experiences on children
's memories. Appl Cogn Psychol 2015;29:791–801.
. Salmon K, Reese E. The benefits of reminiscing
with young children
. Curr Dir Psychol Sci 2016;25:233–8.
. Sheingold K, Tenney YJ. Memory
for a salient childhood event. In: Neisser U, editor. Memory
observed: remembering in natural contexts. New York: W. H. Freeman and Company, 1982. pp. 201–12.
. Stinson JN, Kavanagh T, Yamada J, Gill N, Stevens B. Systematic review of the psychometric properties, interpretability and feasibility of self-report pain intensity measures for use in clinical trials in children
and adolescents. PAIN 2006;125:143–57.
. Taddio A, Chambers CT, Halperin SA, Ipp M, Lockett D, Rieder MJ, Shah V. Inadequate pain management during routine childhood immunizations: the nerve of it. Clin Ther 2009;31(suppl 2):S152–67.
. von Baeyer CL, Marche TA, Rocha EM, Salmon K. Children
for pain: overview and implications for practice. J Pain 2004;5:241–9.
. Walker LS, Levy RL, Whitehead WE. Validation of a measure of protective parent responses to children
's pain. Clin J Pain 2006;22:712–6.
. Walker LS, Williams SE, Smith CA, Garber J, Van Slyke DA, Lipani TA. Parent attention versus distraction: impact on symptom complaints by children
with and without chronic functional abdominal pain. PAIN 2006;122:43–52.