Racial Disparity in Glycemic Outcome During Real World Use of an Advanced Hybrid Closed Loop System (AHCL) Among Youth with Type 1 Diabetes (T1D)
Description
Introduction. African-American (B) youth with Type 1 Diabetes consistently have higher A1c than White (W) patients, even when using open-loop insulin pumps. AHCL technology utilizes glucose sensor data to guide automatic adjustments of insulin by the pump throughout the day to mitigate hyper and hypoglycemia. Potentially AHCL could reduce or eliminate disparity in glycemic outcome in high-risk patient populations Methods and Results. We performed a retrospective chart review of clinic patients who were using Tandem Control IQ (CIQ) AHCL system. Patients were trained in AHCL use by the manufacturer’s representative and then followed routinely in diabetes clinic at the Children’s Hospital of New Orleans (n=62). Data was excerpted from patient eMR and t-connect data bases. Patients were included in the analysis if they had used CIQ for more than 120 days and had officially self-identified in the eMR as either non-Hispanic White (W) or non-Hispanic Black/African-American (B) and had paired data for HbA1c and glucose sensor data at the last clinic visit. Sensor/system data stats were assessed for the 30 days prior to clinic visit. Eight patients self-identified as B and Twenty-nine as W. There was a statistical and clinically significant reduction in A1c (-1.2%±1.6, p=0.0007) for W patients but not B (-0.09±1.6, p=NS) with AHCL use. B patients had higher A1c and sensor Mean Blood Glucose (MBG) with lower percent time in use (TIU) during AHCL mode than W patients. The percent insulin bolus, basal, units of insulin per kg per day, percent glucoses <70 mg>/dL were not different between groups. Overall, latest A1c was significantly correlated with sMBG (r=0.75, p<0.0001), % Time in AHCL Use (r=-0.38, p<0.02) and % Time in glucose target range (r=-0.36, p=0.034). The influence of sMBG and race on the latest A1c was examined in a multiple variable model R2 =0.61, p<0.0001 both race (p=0.0046) and sMBG (p<0.0001) were statistically significant. B patients tended to have HbA1c than W at any given level of MBG. Conclusions. Under real world training and clinic follow up B patients did not have improved HbA1c while W patients had clinically and statistically significant improvement. Percent time spent in CIQ mode was the salient difference between groups with W 87.6% time in use vs 60.1% for B patients. There was no difference in percent time in hypoglycemic range (BG<70 mg/dL) between groups. HbA1c over estimates sensor mean blood glucose of B patients compared to W patients.
Racial Disparity in Glycemic Outcome During Real World Use of an Advanced Hybrid Closed Loop System (AHCL) Among Youth with Type 1 Diabetes (T1D)
Introduction. African-American (B) youth with Type 1 Diabetes consistently have higher A1c than White (W) patients, even when using open-loop insulin pumps. AHCL technology utilizes glucose sensor data to guide automatic adjustments of insulin by the pump throughout the day to mitigate hyper and hypoglycemia. Potentially AHCL could reduce or eliminate disparity in glycemic outcome in high-risk patient populations Methods and Results. We performed a retrospective chart review of clinic patients who were using Tandem Control IQ (CIQ) AHCL system. Patients were trained in AHCL use by the manufacturer’s representative and then followed routinely in diabetes clinic at the Children’s Hospital of New Orleans (n=62). Data was excerpted from patient eMR and t-connect data bases. Patients were included in the analysis if they had used CIQ for more than 120 days and had officially self-identified in the eMR as either non-Hispanic White (W) or non-Hispanic Black/African-American (B) and had paired data for HbA1c and glucose sensor data at the last clinic visit. Sensor/system data stats were assessed for the 30 days prior to clinic visit. Eight patients self-identified as B and Twenty-nine as W. There was a statistical and clinically significant reduction in A1c (-1.2%±1.6, p=0.0007) for W patients but not B (-0.09±1.6, p=NS) with AHCL use. B patients had higher A1c and sensor Mean Blood Glucose (MBG) with lower percent time in use (TIU) during AHCL mode than W patients. The percent insulin bolus, basal, units of insulin per kg per day, percent glucoses <70 mg>/dL were not different between groups. Overall, latest A1c was significantly correlated with sMBG (r=0.75, p<0.0001), % Time in AHCL Use (r=-0.38, p<0.02) and % Time in glucose target range (r=-0.36, p=0.034). The influence of sMBG and race on the latest A1c was examined in a multiple variable model R2 =0.61, p<0.0001 both race (p=0.0046) and sMBG (p<0.0001) were statistically significant. B patients tended to have HbA1c than W at any given level of MBG. Conclusions. Under real world training and clinic follow up B patients did not have improved HbA1c while W patients had clinically and statistically significant improvement. Percent time spent in CIQ mode was the salient difference between groups with W 87.6% time in use vs 60.1% for B patients. There was no difference in percent time in hypoglycemic range (BG<70 mg/dL) between groups. HbA1c over estimates sensor mean blood glucose of B patients compared to W patients.