A randomised controlled provider‐blinded trial of community health workers in sickle cell anaemia: effects on haematologic variables and hydroxyurea adherence. (16th November 2021)

• Record Type: Journal Article
• Authors: Smith, Wally R.; McClish, Donna K.; Lottenberg, Richard; Sisler, India Y.; Sop, Daniel; Johnson, Shirley; Villella, Anthony; Liles, Darla; Yang, Elizabeth; Chen, Ian
• Is Part Of: British journal of haematology. Volume 196:Number 1(2022)
• Page Start: 193
• Page End: 203
• Publication Date: 2021-11-16
• Subjects: Hydroxyurea -- sickle cell disease -- adherence -- case management; Hematology -- Periodicals; Blood -- Diseases -- Periodicals; 616.15
• DOI: 10.1111/bjh.17952
• Languages: English
• ISSNs: 0007-1048

Abstract

Summary: Hydroxyurea (hydroxycarbamide) (HU) for sickle cell anaemia (SCA) is underutilised. Case management is an evidence‐based health management strategy and in this regard patient navigators (PNs) may provide case management for SCA. We hypothesised that HU‐eligible patients exposed to PNs would have improved indicators of starting HU and HU adherence. We randomised 224 HU‐eligible SCA adults into the Start Healing in Patients with Hydroxyurea (SHIP‐HU) Trial. All patients received care from trained physicians using standardised HU prescribing protocols. Patients in the Experimental arm received case management and education from PNs through multiple contacts. All other patients were regarded as the Control arm and received specialty care alone. Study physicians were blinded to the study arms and did not interact with PNs. At baseline, 6 and 12 months we assessed and compared laboratory parameters and HU adherence indicators. Experimental patients had higher 6‐month mean fetal haemoglobin (HbF) levels than controls. But at 12 months, mean HbF was similar, as were white blood cell count, absolute neutrophil count, total haemoglobin, platelet count and mean corpuscular volume. At 12 months there were fewer experimental patients missing HU doses than controls (mean 1·8 vs. 4·5, P = 0·0098), and more recent HU prescriptions filled than for controls (mean 53·8 vs. 92 days, median 27·5 vs. 62 days, P = 0·0082). Mean HU doses were largely similar. We detected behavioural improvements in HU adherence but no haematological improvements by adding PNs to specialty care.