Daily changes in pain, mood and physical function in youth hospitalized for sickle cell disease pain changes in pain, and physical function in

BACkgRounD: Youth with sickle cell disease (SCD) are commonly hospitalized for of painful vaso-occlusive (VOE). However, limited data are available concerning the course of hospitaliza- tion for these children and adolescents and, in particular, whether daily changes occur in pain, emotional status and physical function. oBJeCTiveS: To characterize changes in daily pain intensity, physical function and mood over the course of hospitalization, and to determine whether specific clinical characteristics were associated with these changes. MeThoDS: Daily ratings of pain (0 to 10 numerical rating scale) and mood (Positive and Negative Affect Scale for Children) were completed by 25 youth (11 to 20 years of age) with SCD over a total of 152 days (mean [± SD] = 6.7±5.6 days) of hospitalization. Trained raters determined each youth’s daily physical function. ReSuLTS: Linear mixed modelling was used to examine changes in pain, mood and physical function during hospital stay. The rate of change over the course of hospitalization was significant for reductions in pain inten- sity (P<0.001) and improvements in physical (motor) function (P=0.001). Positive affect over time was significantly associated with subjects’ physi- cal function scores (B 0.24 [95% CI 0.12 to 0.35]) but not with their pain scores. In contrast, negative affect was positively associated with pain and inversely associated with physical function scores (B 1.58 [95% CI 0.23 to 2.93]). ConCLuSionS: The results of the present study demonstrated that children made daily improvements in physical function and pain over hospitalization for VOE. Mood was related to changes in pain and physical recovery. Assessment of physical function and mood during hospital- ization may help guide strategies to better understand the pain experience in youth with SCD hospitalized with VOE. during hospitalization (12). This study used three simple, single-item questions regarding sleep, food intake and activity levels obtained by child self-report to provide assessment of function. Additional investi-gations of physical recovery among youth with acute SCD pain using more comprehensive measures of daily function and multiple reporters (eg, parents, nurses) in the hospital setting may expand on these findings.The goal of the present study was to investigate concurrent daily changes in pain, mood and physical function in children hospitalized for VOE. We focused on expanding knowledge of physical function by using a broad measure that relies on observations of the patient’s functional independence. Specifically, we assessed what aspects of function change during hospitalization, whether pain intensity, mood and functioning change during hospitalization, the relationship between improvements in physical function and improvements in self-reported pain and mood, and whether individual characteristics were associated with changes in pain intensity, function or mood throughout hospitalization.We hypothesized that pain intensity would decrease, physical function would increase, and positive mood would increase and negative mood decrease over days of hospitalization. We also hypothesized that opioid use, length of hospital stay and prehospital physical functioning would be associated with physical functioning over hospitalization. Specifically, we hypothesized that higher opiate doses would be associated with lower physical functioning and that a shorter length of stay would be associated with improved physical functioning. We also expected positive associations between prehospital physical functioning and physical functioning during hospitalization. Finally, we hypothesized that mood would be associated with changes in pain intensity and function, with more positive affect and less negative affect associated with less pain and better functioning. FIM motor examined using linear mixed-effect models examine changes over hospitalization. walking climbing stairs statistically significant improvement during hospitalization (coefficient

during hospitalization (12).This study used three simple, single-item questions regarding sleep, food intake and activity levels obtained by child self-report to provide assessment of function.Additional investigations of physical recovery among youth with acute SCD pain using more comprehensive measures of daily function and multiple reporters (eg, parents, nurses) in the hospital setting may expand on these findings.
The goal of the present study was to investigate concurrent daily changes in pain, mood and physical function in children hospitalized for VOE.We focused on expanding knowledge of physical function by using a broad measure that relies on observations of the patient's functional independence.Specifically, we assessed what aspects of function change during hospitalization, whether pain intensity, mood and functioning change during hospitalization, the relationship between improvements in physical function and improvements in self-reported pain and mood, and whether individual characteristics were associated with changes in pain intensity, function or mood throughout hospitalization.
We hypothesized that pain intensity would decrease, physical function would increase, and positive mood would increase and negative mood decrease over days of hospitalization.We also hypothesized that opioid use, length of hospital stay and prehospital physical functioning would be associated with physical functioning over hospitalization.Specifically, we hypothesized that higher opiate doses would be associated with lower physical functioning and that a shorter length of stay would be associated with improved physical functioning.We also expected positive associations between prehospital physical functioning and physical functioning during hospitalization.Finally, we hypothesized that mood would be associated with changes in pain intensity and function, with more positive affect and less negative affect associated with less pain and better functioning.

Procedures
Children and adolescents (ranging in age from seven to 21 years) with SCD hospitalized for pain consistent with a VOE and with the cognitive ability to report pain intensity on a 0 to 10 numerical rating scale (NRS) were eligible for inclusion in the present study.Patients were excluded if they had a primary diagnosis other than VOE or had concurrent acute chest syndrome (ACS).The study was approved by the institutional review boards at two participating hospitals in the northeastern United States.Patients were approached for consent either on hospitalization or during routine clinical visits in anticipation of future hospitalization.
For subjects younger than 18 years of age, signed consent and Health Insurance Portability and Accountability Act (HIPPA) authorization was obtained from the parent or guardian and the subject gave written assent; subjects 18 years of age or older gave signed consent and HIPPA authorization.Following informed consent/assent and HIPPA authorization, demographic information was obtained from participants' medical records, and parent and patient baseline measures of function, pain intensity and mood were recorded.On each subsequent day of hospitalization, daily measures of pain, function and mood were recorded.The day of transition to oral opioids (defined as taking any oral opioid) and day of discharge were obtained from medical charts.

Participants
Thirty-three patients were screened for study enrollment.Three parents declined to have their children participate and the parents of three additional patients were unavailable to consent.The rate of participation was 84.6%.One patient enrolled in the study but was not hospitalized with VOE during the study period; thus, 26 patients completed data collection.One patient (seven years of age) was excluded from the analyses due to outlying age compared with the rest of the sample.Therefore, the final sample included 25 participants, 11 to 20 years of age (mean [± SD] age 16.6±2.4years).

Background and prehospital variables
Demographics: Demographic information (age, sex, race, sickle cell genotype, previous hospitalizations, emergency department and clinic visits for pain) was obtained through parent report and medical record review.Duration of pain and opioid use before hospitalization were recorded.Prehospital physical functioning: On admission, youth completed the self-report version of the Child Activity Limitations Interview (CALI-21), a 21-item assessment of children's functional limitations due to pain over the previous four weeks (13).The CALI-21 showed good internal consistency, high cross-informant reliability and good construct validity in a sample of 188 children and adolescents (eight to 18 years of age) with chronic pain (13).On the CALI, children are asked to rate how difficult or bothersome routine activities (ie, going to school, gym, playing with friends, housework) are because of pain, on a 0 to 4 Likert scale.Scores are summed and range from 0 to 84, with higher scores indicating more functional impairment due to pain.The child-report scores for this measure provided an assessment of prehospital physical functioning.

Daily hospital assessment variables
Pain intensity: Self-report patient scores on a standard 0 to 10 NRS were recorded by the clinical nurse every 4 h during the patient's hospitalization, unless the patient was sleeping.These were verbal reports of pain recorded in the medical record.A mean pain score was calculated for each hospital day.A mean (± SD) of 5.3±1.2pain scores were recorded daily for each participant.The verbally administered NRS is the most commonly used pain scale in clinical practice and has recently been shown to be a valid assessment tool in children and adolescents with acute pain (14,15).Function: Due to the lack of other available measures of daily physical function, a validated observational and interview measure of functional activity used previously in youth in the inpatient rehabilitation setting was chosen.Daily physical function was, therefore, assessed using the Functional Independence Measure (FIM) (16).Reliability and validity has been established in adolescents 10 to 24 years of age (17), and construct validity has been demonstrated, with FIM ratings yielding significant correlations with clinician ratings of functional status (18).The FIM consists of 18 items that assess function in two domains: motor function (11 items) and cognitive function (seven items) (19).These items include assessments of self-care activities, locomotion, transfers, memory and comprehension.Each of the 18 items is given a score of between 1 and 7, in which 1 indicates that the patient needs 'total assistance' to perform the particular activity and 7 indicates that the patient has 'complete independence' with that activity.Therefore, total scores range between 7 (functional dependence) and 126 (functional independence) (19).The FIM has been validated as both an observational and self-report measure.
For the purpose of the present study, the FIM was administered daily by a trained research assistant.This individual completed training on the FIM and passed a certifying examination for FIM administration.The research assistant completed the FIM based on observations and interviews with the primary nurse and the patient to determine a composite rating for each item.Composite ratings are acceptable as per the FIM instruction manual (19).When there was a discrepancy between staff and patient report, a conservative rule was applied to record the highest level of function for that item.Mood: The Positive and Negative Affect Schedule -Child Version (PANAS-C) is a 27-item (12 positive items, 15 negative items) assessment of mood that was completed by child self-report on each day of hospitalization (20).The PANAS-C was developed and validated in a sample of children in grades four to eight, and demonstrates good convergent and discriminant validity (20).The PANAS-C asks the child to rate the degree to which different words (ie, sad, active, joyful) describe how they feel on a scale of 1 (very slightly or not at all) to 5 (extremely) (20).This scale yields both a positive affect and negative affect score.opioid use: Opioid use was calculated for each 24 h period in intravenous morphine equivalents using a standard conversion formula (21).

Statistical analysis
The demographic characteristics of the sample were described using summary statistics.Means and SDs were used for continuous variables, and categorical variables were summarized using frequencies and proportions.Within-patient changes in pain intensity, physical functioning, cognitive functioning and PANAS-C scores between first and final hospital days were examined using paired t tests.The associations among changes in pain, functioning and mood were assessed using Pearson's correlation coefficients.
The 25 participants were assessed daily for pain intensity, functioning and mood.Days assessed varied from three to 14, generating a total of 152 data points for each measure.To account for the correlation among repeated assessments within participants, linear mixed models were used to examine how pain, functioning and mood varied over hospital stay.Changes of individual items on the FIM over hospital stay were also explored using linear mixed-effects models to identify specific areas of function (eg, gross motor function) in which patients experienced improvements.Similar mixed-effects models were adapted to examine how pain was associated with functioning and whether mood modified such a relationship.Linear mixed models were also used to examine associations among PANAS-C scores, opiate use, prehospital physical functioning and changes in pain and motor functioning.Significance of parameter estimates was assessed using CIs based on simulations (22).
The sample size consideration was mainly based on anticipated reduction in pain intensity during hospital stay.Using a two-sided onesample t test, a sample of 19 participants could provide 90% power to detect a mean within-person reduction of 2.0 with an estimated SD of 2.5 at significance level of 0.05.To account for potential attrition, 25 patients were recruited at baseline.All data analyses were conducted using R statistical software version 2.13.0 (23).

Descriptive statistics
Sample characteristics are summarized in Table 1.Study participants had a mean age of 16.6±2.4years.The majority were female (80%), African-American (76%) and had a diagnosis of Hemoglobin SS (68%).Before their admission, children had experienced a mean of 2.64±3.8days (range 0 to 15 days) of pain.The mean hospital length of stay was 6.7±5.6 days (range two to 28 days).
At admission, participants presented with moderate impairment in motor functioning (mean score 54±12).Minimal impairment in cognitive functioning was reported, with all participants scoring at or above 34 (out of 35 possible) on this subscale.The mean total opiate use during hospitalization was 0.05±0.02mg/kg/h morphine equivalents (range 0 mg/kg/h to 0.33 mg/kg/h).Participants transitioned from intravenous to oral medication in a mean of 6±5.3 days (range two to 28 days).
Values for pain intensity, physical and cognitive functioning, and PANAS scores obtained on the first and final hospital study days are presented in Table 2. Paired t tests were conducted to compare within-subject scores from first to final day of hospitalization.Mean pain intensity scores were significantly lower on the final day compared with the first day of hospitalization.The mean within-subject reduction of pain was −2.3±2.6 (P<0.001).Similarly, total function and motor function both significantly increased on the final day compared with the first day of hospitalization.The mean within-subject increase in the FIM total score was 11.8±15.7 (P<0.001).The mean within-subject increase in the FIM motor scale was 11.7±15.8(P=0.001).Because there was almost no change in FIM cognitive scores, most of the changes in total FIM were driven by the changes in FIM motor scale.Positive affect significantly increased (P<0.02) but negative affect was not significantly different between the first and final day of hospitalization.

Daily change in pain intensity, affect and function over hospitalization
Linear mixed-effect models with both random intercept and random slope and a fixed effect for hospital day (time) were used to assess the rate of change in pain score, physical and cognitive function, and positive and negative affect across hospital stay.Unlike the analysis comparing the outcome measures between first and final hospital days, these analyses aimed to assess the rate of change over time for each measure (Table 3).Results revealed a significant fixed effect of hospital day on pain scores, indicating a significant rate of decrease in pain intensity scores over hospital stay.The estimate suggested that, on average, patients experienced a reduction of 0.5 points on the pain scale with each additional hospital day (B −0.51 [95% CI −0.75 to −0.27]).Total function increased over time but the rate of change was not statistically significant  Individual FIM motor items were examined using linear mixed-effect models to examine changes over hospitalization.Only walking and climbing stairs showed statistically significant improvement during hospitalization (coefficient estimates 0.36 and 0.30, respectively).

Associations among daily pain, function and opiate use
In mixed-effects regression models adjusting for baseline physical functioning on the CALI, higher opiate use was significantly associated with higher pain scores (B 18.68 [95% CI 13 to 25]).However, functional scores were not significantly associated with opiate use.Baseline physical functioning on the CALI was not associated with pain or function during hospitalization.

Associations among pain intensity, function and mood
The longitudinal relationships among pain, function and mood were examined in linear mixed models.Increased pain was not associated with decreased functioning (B -1.07 [95% CI -2.61 to 0.41]).Positive affect, however, was significantly associated with functional scores (B 0.24 [95% CI 0.12 to 0.35]) but not with pain scores.When both pain and positive affect were used as predictors for functioning, pain intensity remained nonsignificant (B -0.19 [95% CI -1.78 to 1.40]) and positive affect remained significant (B 0.44 [95% CI 0.17 to 0.75]).No significant effect modification was detected between pain and positive affect.Unlike positive affect, negative affect was significantly associated with pain scores (B 1.58 [95% CI 0.23 to 2.93]) but not motor function.

DiSCuSSion
Children with SCD experience recurrent episodes of severe, unrelenting pain, frequently requiring inpatient hospital stays.Little is known about the physical recovery or mood of youth hospitalized for VOE, although these are important components of their pain experience.Assessment of functional status during hospitalization may help to guide strategies to address specific functional impairments and better understand the pain experience in children with SCD (12).Attention to mood may lead to better understanding of the factors that lead to and prolong hospitalizations in this patient group.
As observed in our preliminary study of hospitalized youth with VOE, on the first day of hospitalization, youth reported moderate to severe pain and impaired mood/affect, and demonstrated moderate limitations in motor functioning.As hypothesized, we found that youth experienced reductions in pain and improvements in motor functioning over the course of hospitalization.Specifically, we found positive changes in the ability to walk and climb stairs; however, other aspects of motor function, including eating, bathing, dressing/undressing, and toileting/bladder control did not change during hospitalization.Moreover, cognitive aspects of functioning (eg, communication) were in the normal range (not impaired) at the beginning of hospitalization and remained stable over the subsequent hospital days.There was an improvement in positive affect but no change in negative affect from the beginning of hospitalization to the end.
Our findings contribute to the scant available literature on pain, physical function and mood among hospitalized children with SCD.The results partially contrast with those of Jacob et al (12), who found no change in activity level, sleep or eating over hospitalization using daily, global self-report scores obtained from children.The findings may differ due to our use of a broader measure of functional independence, which is more detailed and includes more specific functional activities than the measure used in the study by Jacob et al.It is also possible that caregivers can observe vigorous physical activities more easily than other aspects of motor function, resulting in differing reports compared with children's self-reports.Ideally, assessment of functional status would integrate both child and caregiver reports (24).
Despite the improvement in both motor function and pain during hospitalization, these outcomes were not significantly correlated.Previous studies have demonstrated the lack of direct correspondence between disability and pain, and that behavioural, cognitive and affective factors play a clear role in the extent of functional impairment experienced (25).In SCD in particular, clinicians have noted discrepancies between patient behaviour and pain score (25)(26)(27)(28), which has led clinicians to doubt the reliability of self-reported pain scores in this population (23,27).This often leads to concerns regarding drug-seeking behaviour and opioid addiction (29).These findings demonstrate the importance of evaluating functional outcomes in addition to pain intensity in the acute setting, and also highlight the need for clinicians to understand that pain report and functional status may not correlate.
We also found that positive affect was associated with changes in motor function across hospitalization in our sample, while negative affect was only associated with changes in pain.Previous studies have demonstrated associations between mood and both pain and function (29)(30)(31)(32).In particular, depression has been strongly associated with functional disability in children with chronic pain (29)(30)(31).The complex associations that we found between pain, function and positive and negative mood suggests that in the acute setting, day-to-day changes in mood may play an important role in recovery.This finding demonstrates the need to consider emotional status in the care of children and adolescents hospitalized for VOE.
We did not demonstrate a relationship between prehospital functioning and daily physical functioning during hospitalization in the present study.The CALI evaluates perceived difficulty in physical function due to pain over the previous month in the context of home and school activities.Although we expected that youth with persistent impairments in their physical function related to their SCD pain may demonstrate even more reduced physical functioning during hospitalization for acute VOE, it is likely that the heterogeneity in the pain experience of youth with SCD lessened the relationship between these two functional measures.Moreover, the individual factors that determine physical functioning related to ongoing or persistent SCD pain at home may be very different from factors that determine inhospital functional status in the context of acute vaso-occlusive pain.These important issues are in need of further study to understand the relationship between function in different settings and for different types and intensities of pain.
As the PedIMPPACT group noted, there are no validated instruments for assessing physical recovery or function in youth with acute pain (5).This is in contrast to functional assessment in chronic pain, for which validated measures for children are available (32)(33)(34)(35).In the acute inpatient setting, physical recovery, or the return to the child's pre-illness level of function, includes aspects of the child's motor function (eg, ambulation), as well as other measurements of functional status such as oral intake or swallowing.We chose to use a measure of functional abilities in the present study, although not validated in this specific clinical setting, due to the lack of other available measures and the inclusion of many relevant items of function.In the future, the evaluation of physical recovery of patients hospitalized with VOE will require an instrument that is sensitive to change over the typical two to seven day hospitalization and that also incorporates the patient's perspective.Our preliminary findings may help inform the development of such a measure by identifying key areas of function that change over the course of hospitalization for VOE.
Our study findings should be interpreted in light of several limitations.The present study used a small sample of patients who tended to be older (mean age 16.6 years), and stayed in the hospital longer (6.7 days) than a previous retrospective evaluation performed at our centre (8).This suggests that patients included in the study were more likely those who experienced more severe episodes consistent with longer hospitalization; however, given that pain events in this patient group are complex, it is important to note that improvements in physical function could be measured in our study sample.Patients who are younger and with less severe disease might show even more improvement in function over the course of hospitalization.We were limited in the examination of additional clinical questions about subsets of patients given our small sample size.However, our primary analyses were well powered because they relied on our daily observations, producing a robust dataset to explore daily changes in pain, mood and function, even in this small sample.

SuMMARy
Our preliminary findings demonstrate that pain, mood and physical function change over the course of hospitalization for VOE in children and adolescents with SCD.The lack of correlation between pain intensity and functional recovery in these findings highlight the importance of measuring functional outcomes in this patient group to further understand the impact of VOE.Further work should focus on the development and validation of simple, clinically useful measures to evaluate functional recovery in hospitalized children.Use of these measures, along with attention to positive and negative mood states, will further contribute to the understanding of the complex issues associated with VOE in youth with SCD.DiSCLoSuRe: FIM is a trademark of Uniform Data System for Medical Rehabilitation, a division of UB Foundation Activities, Inc.The use of the FIM instrument to collect data for the present research study was authorized and conducted in accordance with the terms of a special purpose license granted to Licensee by Uniform Data System for Medical Rehabilitation a division of UB Foundation Activities, Inc (UDSMR).The patient data collected during the course of the present research study has not been processed by UDSMR.No implication is intended that such data has been or will be subjected to UDSMR's standard data processing procedures or that it is otherwise comparable with data processed by UDSMR.

FunDing:
Dr Zempsky is supported by the Patrick and Catherine Weldon Donaghue Medical Research Foundation and National Institute of Health grant 1 K23 HL090832-01A1.

Table 1 Demographic and clinical characteristics of the sample (n=25)
* Data were missing for three subjects; † Data were missing for four subjects