Caregiver and Provider Reports of Symptom Severity, Functionality, and Quality of Life Lack Congruence with Patient Reports in Pediatric Sickle Cell Disease

Background:

Pain is the most common symptom of sickle cell disease (SCD), experienced up to 56% of days in patients over 16 years old. SCD pain begins in infancy and affects the patient’s functionality and quality of life. Accurate pain assessment is challenging yet crucial to lifelong management. In children, there are additional barriers to symptom communication. The ability to understand and relay pain symptoms varies depending on age, development, and the presence of additional neurologic sequelae associated with SCD. To overcome these barriers caregivers and medical providers often serve as surrogate symptom reporters for pediatric patients. This leaves room for misunderstanding the true nature and severity of the pain and inappropriate treatment. Studies in pediatric oncology found suboptimal concordance of both caregiver and provider symptom reports with those of the child, but this has not been studied in pediatric SCD.

Methods:

We performed a single site non-randomized cross-sectional study, enrolling patients with SCD age 10 years and older as well as a caregiver and medical provider associated with the visit. Eligible patients completed a visual analogue scale rating, a front and back 2-dimensional body image that can be shaded to indicate areas affected by pain, Lansky Play-Performance Scale (LPP), and validated patient reported outcomes questions (PROMIS Pain interference, PedsQL General Well-Being and Health, Ped-PRO-CTCAE). Patient responses were analyzed for inter-rater reliability compared to caregiver and provider responses.

Results:

We enrolled 29 patient-caregiver dyads and 28 patient-provider dyads. The mean age of patients enrolled was 16 years and 48% were females. Seventy-four percent of caregivers were female. Eighty eight percent of providers were advanced practice providers, and the remaining were evenly distributed between physicians and psychologists. Not all patients had associated reports from both caregivers and providers. There were 23 complete triads, 6 caregiver only dyads, and 5 provider only dyads. Regarding correlation between reports from the patient and each dyad partner, the caregiver-patient agreement on LPP was fair (κ=0.25) and provider-patient agreement was negligible (κ=0.08). Both caregiver and provider agreement on VAS were negligible ( κ=0.08 and κ=0.10, respectively). All PRO tool agreements exhibited negligible Cohen’s kappa values of < 0.2. On the CTCAE/Ped-PRO-CTCAE, the percentage of providers who reported no symptom (e.g. pain, nausea, fatigue, neuropathy) was consistently higher than the percentage of patients reporting no symptom (median of 78.6% vs. median of 17.9%). Caregiver and provider VAS scores and LPP reports were not significantly different from patient reports; however, parents tended to report higher pain severity and lower functionality while providers tended to report lower pain severity and higher functionality. For PedsQL and PROMIS, there were comparable numbers of patients and caregivers reporting the highest quality of life with a range of 0-3.

Conclusions:

We found that when compared to patient reports, caregiver and provider reports of symptom severity, general functionality, and quality of life exhibited poor agreement. This study adds to data in pediatric oncology highlighting the lack of congruence in caregiver and provider symptom report with the pediatric patient’s report. Attention to the child’s report and improved tools for symptoms assessment are needed to provide children with more agency to communicate their experiences.