Long-term Pain Symptomatology in PICU Survivors Aged 8–18 Years

Abstract

BACKGROUND AND OBJECTIVES

Children discharged from the PICU often experience long-term physical, psychological, social and cognitive challenges, described as postintensive care syndrome. This study fills a critical gap by describing the long-term pain symptoms many children experience using child self-report.

METHODS

Participants in this cross-sectional study were children aged 8 to 18 years (and 1 parent, n = 64 dyads) who were 8 to 24 months post-PICU discharge. Participants completed measures assessing pain, psychosocial function, and treatment utilization. Health information relevant to the PICU admission was obtained from the electronic medical record.

RESULTS

Children were an average age of 14.3 years and 50% female. Average pain intensity in the past month was 3.65 (0–10 numeric rating scale), with 36% of children reporting pain ≥2 to 3 days a week. Youth with higher intensity (≥4) and more frequent pain (≥2–3 days a week) had greater pain catastrophizing, pain-related fear, anxiety, and sleep disturbance than those with less frequent and intense pain. Higher pain frequency was also associated with greater pain-related disability and depressive symptoms. Pain was not associated with any PICU-related variables abstracted from the electronic medical record. Parents reported children frequently accessed pain treatment services postdischarge; however, 37.5% reported the coronavirus disease 2019 pandemic impacted access to recommended services.

CONCLUSIONS

A significant portion of children experience pain post-PICU discharge. For many of these children, pain is frequent and is associated with impairments in psychosocial function. Future prospective research studies can be used to identify risk factors of poor pain outcomes so children can receive targeted interventions.

Three hundred thousand children are admitted to the PICU annually in the United States¹ and steep decreases in PICU mortality highlight progress in treating critically ill and injured children.² Despite these gains, children discharged from the PICU experience significant long-term physical, psychological, social, and cognitive challenges.³ This constellation of symptoms, described as postintensive care syndrome (PICS-p), has received recent attention with a key focus being improved assessment of PICS-p and identification of modifiable factors to improve outcomes.^4–6

Given the severity of illness and injury in children who survive critical care experience, studying pain in children discharged from the PICU is critical and was highlighted as a priority for outcomes research.^7,8 Adult data show 33% to 77% of patients develop chronic ICU pain postdischarge.⁹ There is a key gap in understanding pain trajectories in PICU survivors. Although important work has characterized some physical and cognitive outcomes,^10,11 and psychological sequelae including PTSD,¹² studies examining long-term pain symptomatology in this population, and research on risk and resilience factors associated with pain persistence have not been conducted. Adults with chronic pain onset in childhood have higher rates of disability, psychological problems, unemployment, and opioid misuse than those whose pain started later.^13–15 Moreover, both history of ICU admission and chronic pain increase the risk of suicide and early death.^16–20 Understanding and characterizing pain post-PICU discharge can support the development of targeted interventions for children at highest risk for developing persistent pain.

A critical component of understanding pain in PICU survivors is obtaining self-report data. Data on pain in PICU survivors have often relied on parent or proxy report. There are often differences in child versus parent reports of pain, with the recommendation to obtain child self-report when feasible.^21,22 It is also important to assess how cognitive–affective and behavioral factors (eg, pain catastrophizing, pain-related fear, and parental pain protective behaviors), known to be important in pediatric pain samples,^23–25 are associated with pain-related functioning in children discharged from the PICU. To our knowledge, previous research has characterized more general psychological functioning in PICU survivors but not explored pain-specific cognitive–affective and behavioral factors that may impact health outcomes in this sample.

There is also a gap in understanding how treatments accessed postdischarge relate to pain symptomatology over time. At time of discharge, many children are given recommendations for follow-up treatment. However, what treatments children most commonly access for pain is unknown.

To fill these critical gaps, we conducted a study examining pain and pain-related functioning in children post-PICU discharge using a combination of electronic medical record (EMR) data and child and parent report. We intentionally included children with a number of primary diagnoses across the spectrum of illness and injury severity to fully characterize pain in a diverse post-PICU sample. Aim 1 was to describe long-term pain symptomatology and pain-related function in children and the concordance among parent and child reports. Aim 2 was to examine associations among long-term pain symptomatology and psychosocial function. Aim 3 was to describe treatment utilization for pain postdischarge, as well as the impact of the coronavirus disease 2019 (COVID-19) pandemic on access to recommended treatments.

Methods

This cross-sectional study was conducted at a children’s hospital in the northwestern United States. Study procedures were approved by the institutional review board and participants provided consent or assent before participating. Families of children aged 8 to 18 years who were 8 to 24 months post-PICU discharge were invited to participate.

Four hundred twenty-two potentially eligible patients admitted between May 1, 2019, and May 1, 2021, were identified via the EMR. EMR search criteria specified: Child was 8 to 17 years at time of PICU admission, child is currently alive, and the preferred language in medical record was listed as English. A total of 204 patients were unreachable (eg, invalid contact information; voicemail not returned). A total of 218 parents/caregivers were assessed for interest/screened for eligibility, 15 were ineligible, and 46 declined. An additional 44 parents did not sign the consent form after the e-mail consent link was sent. Children with a previous PICU admission were excluded from analyses (n = 15) as were families (n = 34) without dyadic data (eg, youth self-report was not possible because of cognitive impairment). Final sample was n = 64 dyads. See Supplemental Fig 1 for study flow.

Procedures

Research staff contacted caregivers of potentially eligible children via phone, MyChart message (a secure messaging for health care interactions), and/or e-mail about the study. Both parents/caregivers and children needed to be fluent in English to participate. Parents/caregivers needed to be legal guardians to provide consent. Interested and eligible children/parents provided electronic consent/assent via REDCap²⁶; unique links to complete the secure REDCap surveys were e-mailed.

Measures (See Supplemental Table 6 for Additional Information)

Demographic Questionnaire

Parents reported on sex, race, ethnicity, education level, household income, and child insurance status.

Electronic Medical Record

Information related to PICU admission including age, date of admission, length of hospital stay (LOS), and PICU-related diagnoses.

Primary PICU Diagnosis

Expertly coded by PICU physician from EMR data as 1 of 9 categories (trauma, ingestion, respiratory, hematology/oncology, endocrine, cardiac, neurologic, sepsis/infection, other surgical procedures).

Organ Dysfunction

Pediatric Logistic Organ Dysfunction-2 (PELOD-2) is a quantitative scoring system in children based on assessments of 5 organ systems. Scores of ≥10 are associated with mortality.²⁷ The PELOD-2 maximum score during admission was used in analyses.

Functional Status at PICU Admission

The Functional Status Scale (FSS) assesses 6 domains of functioning (mental status, sensory, communication, motor function, respiratory status, and feeding). FSS was assigned by a PICU attending physician at admission to reflect baseline function before hospitalization.²⁸ Scores 6 to 7 are considered “normal,” with ≥8 categorized as “impaired.”

Children’s Pain Characteristics

Children and parents (proxy report) reported on children’s pain intensity (0–10 numeric rating scale [NRS]),²⁹ frequency (pain days per week), and number of pain locations (via body map) over the past month.

Pain Catastrophizing

Using the Pain Catastrophizing Scale self-report and parent proxy report versions, children reported how much they catastrophize about their own pain and parents reported on how much they catastrophize about their child’s pain.^30–32

Fear of Pain

The Fear of Pain Questionnaire²³ was used to assess fear and avoidance related to pain experiences.

Protective Responses to Pain

The Protect subscale of the Adult Responses to Children’s Symptoms³³ was used to assess parents’ use of pain-protective behaviors via both parent- and youth-report.

Treatment Utilization

Parents reported on treatment utilization for pain since PICU admission. Items assessed if the child had accessed certain treatments (yes/no), frequency of use, and perceived benefit.³⁴ One additional question assessed impact of COVID-19 on access to “recommended medical care/services.”

Anxiety, Depression, Sleep

Youth completed the anxiety, depression, and sleep subscales of the Patient-Reported Outcomes Measurement Information System Pediatric Profile-25.³⁵

Functional Disability

Youth completed the Functional Disability Inventory to assess difficulty engaging in activities because of physical health.³⁶

Data Analysis

Data were analyzed using SPSS v.28. Summary statistics described characteristics of the sample. Means, SDs, and/or medians with interquartile range were used for continuous data, and categorical items were described using frequency statistics. Concordance among child- and parent-report of pain and other variables was assessed using correlations. t tests compared children who had clinically significant pain intensity (≥4 NRS)^37,38 and frequency (≥2–3 days a week), versus those with less intense and frequent pain. Data were considered missing at random.

Results

Participants were 64 child–parent dyads. Participant demographic and descriptive statistics are located in Table 1. Descriptives of pain, disability, and psychosocial functioning are located in Table 2 (child-report) and Supplemental Table 7 (parent-report data). Youth were aged 8 to 18 years (mean, 14.27; SD, 2.28), 50% were female; parents were aged 28 to 61 years (mean, 42.14; SD, 6.07) and were 82.9% biological/adoptive mothers.

TABLE 1.

Sociodemographic and Clinical Characteristics of the Sample (N = 64 Child–Parent Dyads)

Parent/Family Demographics		Youth Demographics
Age, mean (SD)	42.14 (6.07)	Age, mean (SD)	14.27 (2.28)
Sex assigned at birth, n (%)		Sex assigned at birth, n (%)
Male	9 (14.1)	Male	32 (50)
Female	55 (85.9)	Female	32 (50)
Gender identity, n (%)		Gender identity, n (%)
Male	9 (14.1)	Female	28(43.8)
Female	54 (84.4)	Male	32 (50)
Unknown/not reported	1 (1.6)	Nonbinary	1 (1.6)
		“Gender I identify as is not listed.”	2 (3.1)
		Unknown/not reported	1 (1.6)
Parent ethnicity, n (%)		Child ethnicity, n (%)
Hispanic/Latino	6 (9.4)	Hispanic/Latino	8 (12.5)
Non-Hispanic	58 (90.6)	Non-Hispanic	55 (85.9)
		Unknown/not reported	1 (1.6)
Parent race, na (%)		Child race, na (%)
American Indian/Alaska Native	4 (6.3)	American Indian/Alaska Native	4 (6.3)
Native Hawaiian/Pacific Islander	2 (3.1)	Asian American	1 (0.9)
White	61 (95.3)	Native Hawaiian/Pacific Islander	2 (3.1)
Other	1 (1.6)	White	60 (93.8)
		Other	2 (3.1)
Relationship to child, n (%)		PICU diagnosis, n (%)
Biological/adoptive mother	53 (82.9)	Trauma	7 (10.9)
Biological/adoptive father	9 (14.1)	Ingestion	11 (17.2)
Other relative or guardian	1 (1.6)	Respiratory	8 (12.5)
Unknown/not reported	1 (1.6)	Hematology/oncology	10 (15.6)
		Endocrine	15 (23.4)
		Cardiac	5 (7.8)
		Neurologic	3 (4.7)
		Sepsis/infection	3 (4.7)
		Other surgical procedures	2 (3.1)
Parent education, n (%)		Child insurance, n (%)
High school or less	6 (9.4)	Private insurance	37 (57.8)
Vocational school/some college	26 (40.6)	Medicaid/state insurance	23 (35.9)
College	21 (32.8)	Other	4 (6.3)
Advanced degree	11 (17.2)
Parent marital status, n (%)		PICU variables	M (SD)
Married	39 (60.9)	LOS	6.55 (8.47)
Divorced/separated	11 (17.2)	Time since discharge	456 (143)
Remarried	2 (1.8)	D on ventilator	0.8 (2.49)
Never married	2 (3.1)	PELOD-2 score	3.44 (3.83)
Other	4 (6.3)	FSS score	6.36 (1.35)
Unknown/not reported	1 (1.6)
Household income, n (%)
<$25 000	7 (10.9)
<$49 999	12 (18.8)
$50 000–$79 999	10 (15.6)
$80 000–$119 999	24 (21.2)
$120 000–$149 999	7 (10.9)
>$150 000	13 (20.3)
Unknown/not reported	1 (1.6)

^aParticipants were able to select >1 option for race/ethnicity identification.

TABLE 2.

Pain, Disability, and Psychosocial Functioning (Youth-Report)

Characteristic	% or Mean (SD)	Median (IQR)
Pain frequency, past mo, n = 63		1 (0–2)
None	25 (39.1%)
1 d per wk	16 (25%)
2–3 d per wk	17 (26.6%)
4–6 d per wk	2 (3.1%)
Daily	4 (6.3%)
Pain intensity, past month, NRS 0–10, n = 63	2.08 (2.02)	2 (0–4)
0, no pain	22 (34.4%)
1	7 (10.9%)
2	9 (14.1%)
3	7 (10.9%)
4	10 (15.6%)
5	5 (7.8%)
6	1 (1.6%)
7	2 (3.1%)
Pain-related disability (FDI), n = 56	6.95 (10)	2 (0–8.9)
No/minimal disability	45 (70.3%)
Moderate disability	8 (12.5%)
Severe disability	3 (4.7%)
Missing	8 (12.5%)
Pain catastrophizing, n = 56	10.47 (9.80)	9 (2.37–15)
Fear of pain, n = 56	17.42 (18.00)	13.28 (3–25)
Parent pain protectiveness, n = 56	20.53 (13.75)	21 (9.45–28.36)
Depressive symptoms, n = 61	50.60 (10.76)	49.80 (38–60.7)
Anxiety symptoms, n = 60	49.71 (11.10)	52 (36–59.50)
Sleep disturbance, n = 59	56.30 (10.61)	55.80 (48.8–64)

FDI, Functional Disability Inventory; IQR, interquartile range.

Average time since PICU discharge was 456 days (range 244–739). Mean LOS (inclusive of PICU stay) was 6.55 days (SD 8.47). Most frequent diagnoses associated with admissions were: endocrine (n = 15, 23.4%), ingestion (n = 11, 17.2%), hematology/oncology (n = 10, 15.6%) and respiratory (n = 8, 12.5%). Maximum admission PELOD-2 scores ranged from 0 to 24 (mean, 3.44; SD, 3.83), with the majority (93.7%) of participants receiving scores <10. Twenty percent received mechanical ventilation during PICU stay. Preadmission baseline FSS ranged from 6 to 14 (mean, 6.36; SD, 1.35), with 92.2% classified as normal (score 6–7).

Aim 1 described pain and pain-related functioning in children, and compared children and parent-proxy report. Youth reported average pain intensity of 3.65 (SD 4.80), with number of pain locations ranging from 0 to 25 (mean, 3.64; SD, 4.80). On proxy report, parents reported average child pain intensity of 3.55 (SD, 5.66), with number of pain locations ranging from 0 to 30 (mean, 3.55; SD, 5.66). Youth- and parent-proxy reports of pain were correlated in terms of both pain frequency (bivariate correlation coefficient [r][62] = 0.52, P < .001) and pain intensity (r[62] = 0.67, P < .001) in the past month. Furthermore, child-report of parent pain protectiveness was correlated with parent’s own report of their engagement in pain protective behaviors (r[54] = 0.50, P < .001).

Thirty-six percent of children reported pain frequency ≥2 to 3 days a week, with 6.3% of the sample reporting daily pain. Nearly one-third of (28.6%) children rated average pain in the past month as ≥4 (0–10 NRS). r revealed that neither pain frequency, intensity, nor number of pain locations was associated with any PICU-related variables (days since discharge, LOS, mechanical ventilation, PELOD-2 score, or FSS) (Table 3).

TABLE 3.

Spearman Correlations Between Youth Reported Pain Outcomes and PICU Admission Variables

	Pain Frequency	Pain Intensity	FDI	FOPQ	PCS, Child-Report	PCS, Parent-Report	Pain Locations (Child-Report)	Pain Locations (Parent-Report)	Parent Protective Behaviors (Parent-Report)	Parent Protective Behaviors (Child-Report)
LOS	−0.01	−0.20	0.09	0.02	−0.08	−0.16	−0.13	−0.05	−0.04	0.19
PELOD maximum	0.05	−0.15	0.16	0.10	0.08	0.11	−0.11	0.16	0.02	−0.02
D on ventilator	0.20	0.08	0.03	−0.06	−0.07	−0.10	0.12	0.15	−0.14	−0.09
Baseline FSS score	0.12	0.10	0.17	0.01	0.02	0.17	0.02	0.03	0.19	0.07

FDI, Functional Disability Inventory; FOPQ, Fear of Pain Questionnaire; PCS, Pain Catastrophizing Scale.

Youth-report of pain-related disability varied, with Functional Disability Inventory scores ranging from 0 to 53; 17.2% met the clinical cutoff for moderate or severe pain-related disability. Children with more frequent versus less frequent pain (≥2–3 days a week versus 1 day per week or less; t[54] = −2.40 [−11.76 to −1.03], P = .02) had significantly higher pain-related disability. There was no difference in pain-related disability by children with higher versus lower intensity (4 versus <3; 0–10 NRS) pain.

Aim 2 examined associations among psychosocial functioning and pain and pain-related disability variables. Correlations among pain, pain-related disability, and psychosocial factors are located in Table 4. Patterns of correlations revealed significant associations among pain characteristics (frequency, intensity, number of pain locations) and pain-related disability, pain catastrophizing, anxiety, depression, and sleep disturbances. Pain intensity but not pain frequency was associated with pain-related fear. Number of pain locations was significantly associated with anxiety and sleep disturbance.

TABLE 4.

Bivariate Correlations of Youth-Reports of Pain and Psychosocial Functioning

1. Pain frequency	—
2. Pain intensity	0.80**	—
3. Number of pain locations	0.52**	0.51**	—
4. FDI	0.52**	0.43**	0.25	—
5. PCS	0.31*	0.44**	0.08	0.46**	—
6. FOPQ	0.18	0.32*	0.04	0.44**	0.84**	—
7. Pain protectiveness	−0.02	0.12	−0.01	0.28*	0.36**	0.37**	—
8. Anxiety	0.38**	0.42**	0.35**	0.26	0.18	0.15	0.03	—
9. Depression	0.31*	0.33**	0.20	0.36**	0.35**	0.37**	0.30*	0.61**	—
10. Sleep disturbance	0.44**	0.47**	0.26*	0.40**	0.44**	0.43**	0.11	0.46**	0.51**

FDI, Functional Disability Inventory; FOPQ, Fear of Pain Questionnaire; PCS, Pain Catastrophizing Scale.

^*P < .05;

^**P < .01. —, indicates that the value is not included as the construct would be correlated with itself.

Children with more frequent pain (≥2–3 days a week) had higher pain catastrophizing (t[54] = −2.17, P = .04; 95% confidence interval [−11.02 to −0.42]), greater depressive symptoms (t[59] = −2.08, P = .04; [−11.40 to −0.22]), more anxiety symptoms (t[58] = −2.71, P < .01 [−13.45 to −2.02]), and higher sleep disturbance (t[57] = −3.49, P < .001 [−14.68 to −3.97]), than children with less frequent pain (1 day per week or less). Similarly, children with clinically significant pain intensity (pain ≥4) reported higher pain catastrophizing (t[54] = −3.47, P = .001; 95% confidence interval [−14.78 to −3.96]), more pain-related fear (t[54] = −2.37, P = .02 [−21.81 to −1.89]), more anxiety (t[58] = −2.13, P = .04 [−12.99 to −0.39]), and greater sleep disturbance (t[57] = −3.34, P = .001 [−15.63 to −3.91]) than children with lower pain levels. There were no associations among pain intensity and either pain-related disability or symptoms of depression.

Aim 3 examined parent-report of children’s treatment utilization since discharge and the impact of the COVID-19 pandemic on families’ ability to access recommended services (Table 5). The most commonly accessed treatments for pain were over-the-counter pain medication (52.5%), mental health services targeting pain and function (26.6%), and physical therapy (18.8%). EMR data revealed 18.8% were prescribed an opioid at discharge; parents reported 17.2% of children used the opioid prescription. Perceived helpfulness of interventions accessed was generally high, with the majority of domains rated ≥4 (0–6 NRS). A total of 37.5% of parents reported the COVID-19 pandemic impacted their ability to access recommended services for their child. No difference was found in number of pain treatment services accessed by parents who did (mean, 1.88; SD, 1.45) versus did not (mean, 1.65; SD, 1.20) report that COVID-19 impacted their child’s access to recommended services.

TABLE 5.

Treatment Utilization Since PICU Discharge, N = 62

Pain Treatment Types	Usage Since PICU Discharge
OTC medications for pain, n (%)	32 (52.5)
Frequency, %
Less than once per mo	21.9
1–3 times per mo	56.3
1–3 times a wk	21.9
Most d	0
Perceived helpfulness, scale of 0–6, mean (SD)	4.25 (SD 1.62)
Melatonin, n (%)	29 (45.3)
Frequency, %
Less than once per mo	10.9
1–3 times per mo	6.3
1–3 times a wk	14.1
Most d	14.1
Perceived helpfulness, scale of 0–6, mean (SD)	3.18 (2.07)
Mental health treatment, n (%)	17 (26.6)
Frequency, %
Less than once per mo	17.6
1–3 times per mo	52.9
1–3 times a wk	29.4
Most d	0
Perceived helpfulness, scale of 0–6, mean (SD)	4.33 (1.95)
Physical therapy, n (%)	12 (18.8)
Frequency, %
Less than once per mo	25
1–3 times per mo	41.7
1–3 times a wk	25
Most d	8.3
Perceived helpfulness, scale of 0–6, mean (SD)	4.67 (1.61)
Prescription medication for pain, n (%)	11 (17.2)
Frequency, %
Less than once per mo	18.2
1–3 times per mo	36.4
1–3 times a wk	0
Most d	45.5
Perceived helpfulness, scale of 0–6, mean (SD)	5.00 (1.18)
Weight/strength training, n (%)	10 (15.6)
Frequency, %
Less than once per mo	10
1–3 times per mo	30
1–3 times a wk	60
Most d	0
Perceived helpfulness, scale of 0–6, mean (SD)	3.67 (1.80)
Pool/swimming, n (%)	7 (10.9)
Frequency, %
Less than once per mo	28.6
1–3 times per mo	57.1
1–3 times a wk	14.3
Most d	0
Perceived helpfulness, scale of 0–6, mean (SD)	4.14 (2.41)
Massage, n (%)	4 (6.3)
Frequency, %
Less than once per mo	25
1–3 times per mo	50
1–3 times a wk	25
Most d	0
Perceived helpfulness, scale of 0–6, mean (SD)	4.75 (1.30)
Herbal medicines, n (%)	4 (6.3)
Frequency, %
Less than once per mo	50
1–3 times per mo	25
1–3 times a wk	25
Most d	0
Perceived helpfulness, scale of 0–6, mean (SD)	5.00 (1.73)
Yoga, n (%)	4 (6.3)
Frequency, %
Less than once per mo	0
1–3 times per mo	100
1–3 times a wk	0
Most d	0
Perceived helpfulness, scale of 0–6, mean (SD)	3.50 (2.08)
Chiropractic treatments, n (%)	3 (4.7)
Frequency, %
Less than once per mo	0
1–3 times per mo	66.7
1–3 times a wk	33.3
Most d	0
Perceived helpfulness, scale of 0–6, mean (SD)	4.33 (2.89)
TENS unit, n (%)	1 (1.6)
Frequency, %
Less than once per mo	100
1–3 times per mo	0
1–3 times a wk	0
Most d	0
Perceived helpfulness, scale of 0–6, mean (SD)	0 (0)

OTC, over the counter; TENS, transcutaneous electrical nerve stimulation.

Discussion

Although an increasing amount of research points to long-term sequela after PICU admission including physical, psychological, social, and cognitive challenges,^3,5,6 a dearth of research has characterized long-term pain symptomatology. Knowing pain trajectories of children is critical for fully characterizing functioning over time, identifying supports needed in recovery, and preventing chronic pain across the lifespan. Results of the current study revealed 29% of children endorsed clinically significant (≥4 NRS) pain at long-term follow-up. Thirty-six percent reported pain frequency ≥2 to 3 days a week in the past month. The percentage of children reporting pain 8 to 24 months post-PICU discharge is comparable to rates of chronic pain in adult ICU samples⁹ and chronic pain in children after major surgeries.^39,40 This data highlight pain as an important component of PICS-p and suggest more detailed assessment of pain trajectories post-PICU admission is warranted to identify early risk and resilience factors associated with pain persistence over time.

Findings showing associations among pain variables and psychological function (eg, pain catastrophizing, pain-related fear, anxiety, and depression) highlight how closely pain is related to day-to-day functioning in this sample. Although associations among anxiety, depression, and posttraumatic stress in PICU survivors are documented,^6,41 associations between pain and mental health sequela in post-PICU samples have not been well studied. Of particular importance are associations among pain and both pain catastrophizing (eg, rumination, magnification) and pain-related fear. To our knowledge, this is the first study to assess pain catastrophizing and pain-related fear in a post-PICU sample and examine associations with long-term pain symptomatology. In general pediatric chronic pain samples, pain catastrophizing has been associated with pain intensity and disability, both concurrently and longitudinally.^42–44 Findings suggest screening for fear–avoidance and catastrophizing in children with elevated symptomatology may be important in pain recovery in post-PICU samples. Tools such as the Pediatric Pain Screening Tool, which have been validated for use with acute and chronic pain samples, may be useful for assessing risk at PICU discharge.⁴⁵ Additionally, given the increasing focus on post-PICU outcomes and numbers of PICU follow-up programs,^7,8 providers should consider pain assessment and intervention a priority.

Study findings also showed strong correlations between pain intensity, pain-related outcomes, and sleep disturbances. This is consistent with previous research in pediatric pain, which indicates bidirectional sleep–pain associations.⁴⁶ Although sleep disturbances have previously been reported at elevated rates in children post-PICU⁴⁷ and sleep impacts development, cognitive function, and emotional health⁴⁸ in children post-PICU,^49–52 our study is among the first to suggest potential associations between pain and this important PICS-p morbidity. Consistent with the general pain literature, our study suggests that a multifaceted approach that addresses both sleep and pain may optimize recovery across these health domains.

A key strength of this study is that pain and pain-related symptomatology were assessed using both child self-report and parent-proxy report data. Reliance on parent-report is common in PICU samples and often warranted given the severity of illness and injury children experience. Results from this study show many children can report on their own pain and pain-related symptomatology. Although our results showed good correlation between child- and parent-proxy report of pain, we contend the child perspective offers additional valuable data. Future PICU studies should continue to use child-report of pain and assess how pain changes over the recovery process.

This study also provides important information on treatment accessed for pain post-PICU discharge. Youth use a variety of pain interventions, and results presented highlight generally positive ratings of perceived helpfulness of interventions accessed, particularly nonpharmacologic interventions. It is unknown if children in this sample would have less pain or pain-related disability with better access to treatment, which can be examined in future research. Additionally, studies should examine any discrepancies between provider treatment recommendations and patient utilization.

Some study limitations could have affected findings or generalizability of results. First, data were collected from only n = 64 dyads, and participants were children ≥8 years of age with different PICU diagnoses who received different medical interventions. Although focusing on a specific diagnosis would allow greater breadth of risk factor analysis within the acute phase, our intention was to survey a broad sample akin to the typical case mix of PICU admissions^53,54 and explore pain in the context of PICS-p. It is also recognized that a portion of participants may have had pain before PICU admission. Although FSS scores at baseline were normal for 97.7% of the sample, the FSS does not include pain-specific items and a variety of chronic medical or psychological conditions could contribute to pain risk not captured by FSS. Our study design precluded assessment of pain before PICU admission because pain is not standardly reported in the EMR. Survey items on preadmission symptoms were not included because of concerns regarding recall bias in posthospitalization samples^55,56 Temporal patterns of pain pre-, during, and post-PICU admission can be assessed in future research. This research should also include assessment of pain in nonverbal and young children. A host of tools are available for assessing pain in nonverbal children of a variety of ages.^57–59 An additional limitation is the limited racial/ethnic diversity of the sample. Although study participants reflect the racial and ethnic breakdown of patients seen at this institution, future studies need to recruit larger and more diverse samples to determine if results are generalizable to non-English–speaking samples and those with greater racial/ethnic diversity. Finally, participants in the study were all either hospitalized or in their post-PICU recovery period during the COVID-19 pandemic. Without a comparison group (eg, pre–COVID-19 pandemic sample), we were not able to assess how potential pandemic-related factors impacted pain and related disability.

Conclusions

Data from this study contribute to the dearth of knowledge on long-term pain symptomatology in children discharged after a first PICU admission. Findings highlight that a significant number of children continue to have frequent and impairing pain even up to 2 years postdischarge, indicating pain should be routinely assessed and treated as part of PICU follow-up in the context of PICS-p and chronic illness care. Future work can include development of additional screening protocols for assessing pain, fear of pain, catastrophizing, and pain-related disability for PICU survivors and families. Additionally, prospective longitudinal studies beginning at time of PICU admission are needed to examine how pain changes over time and how use of specific interventions (both during admission and postdischarge) impact pain and disability outcomes. In the interim, as the field of post-PICU follow-up continues to grow,⁸ programs should incorporate assessment and treatment recommendations for pain in standard PICS-p care.