For all results, the “2020” group represents encounters occurring between 3/15/2020 (the approximate day that COVID restrictions were imposed in Dallas County) and 3/14/2021; the “2019” group comprises encounters between 3/15/2019 and 3/14/2020. The demographic composition of the 2020 group is presented in Table 1.
Table 1
Outpatient visit frequency
|
Visits
|
|
0
|
1
|
2
|
≥ 3
|
Total
|
P+
|
Count of patients with each number of total outpatient visits, by year*
|
Commercial
|
|
|
|
|
|
|
2019
|
|
--
|
150
|
254
|
649
|
1053
|
|
|
%
|
--
|
14.2
|
24.1
|
61.6
|
100
|
|
2020
|
|
--
|
252
|
331
|
401
|
984
|
|
|
%
|
--
|
25.6
|
33.6
|
40.8
|
100
|
|
Total
|
--
|
402
|
585
|
1050
|
2037
|
< 0.0001
|
Non-commercial
|
|
|
|
|
|
|
2019
|
|
--
|
139
|
172
|
382
|
693
|
|
|
%
|
--
|
20.0
|
24.8
|
55.1
|
100
|
|
2020
|
|
--
|
150
|
179
|
318
|
647
|
|
|
%
|
--
|
23.2
|
27.7
|
49.1
|
100
|
|
Total
|
--
|
289
|
351
|
700
|
1340
|
0.09
|
Count of patients with each number of office and virtual visits in 2020
|
Office
|
|
|
|
|
|
|
|
Commercial
|
|
262
|
302
|
264
|
156
|
984
|
|
%
|
26.6
|
30.7
|
26.8
|
15.9
|
100
|
|
Non-commercial
|
|
136
|
197
|
181
|
133
|
647
|
|
|
%
|
21.0
|
30.4
|
28.0
|
20.6
|
100
|
|
Total
|
|
398
|
499
|
445
|
289
|
1631
|
0.02
|
Virtual
|
|
|
|
|
Commercial
|
|
363
|
391
|
168
|
62
|
984
|
|
|
%
|
36.9
|
39.7
|
16.5
|
6.6
|
100
|
|
Non-commercial
|
|
240
|
251
|
107
|
49
|
647
|
|
|
%
|
37.1
|
38.8
|
16.5
|
7.6
|
100
|
|
Total
|
|
603
|
642
|
275
|
95
|
1631
|
NS
|
+ P values by Fisher Exact Tests |
*Patients were included in the analysis for a given year only if they had at least one outpatient visit. |
In response to the pandemic, all clinic visits were suspended on 3/16/2020 and a virtual visit platform was quickly put in place starting on 4/1/2020. Our operations gradually transitioned back to clinic visits starting on 5/1/2020 and gradually ramped up.
In 2020, the number of outpatient visits (including both clinic and virtual visits) per patient decreased markedly for those with commercial insurance but there was not a significant decrease per patient with non-commercial insurance (Table 2). Patients with commercial insurance had fewer office visits per patient than those with non-commercial insurance. However, there was no difference in utilization of the virtual visit platform in patients based on insurance status.
Table 2
Count of patients with each number of admissions, by year
|
Commercial
|
Admissions
|
|
0
|
1
|
>2
|
Total
|
P
|
2019
|
|
992
|
47
|
14
|
1053
|
|
%
|
94.2
|
4.5
|
1.3
|
100
|
|
2020
|
|
916
|
59
|
9
|
984
|
|
%
|
93.1
|
6.0
|
0.9
|
100
|
|
Total
|
|
1908
|
106
|
23
|
2037
|
NS
|
Non-Commercial
|
581
|
82
|
30
|
693
|
|
2019
|
|
|
|
83.8
|
11.8
|
4.3
|
100
|
|
2020
|
|
543
|
75
|
29
|
647
|
|
|
83.9
|
11.6
|
4.5
|
100
|
|
Total
|
|
1124
|
157
|
59
|
1340
|
NS
|
P values by Fisher Exact Tests.
P<0.0001 commercial vs non-commercial.
|
There was no change in hospitalization rates from 2019 to 2020 in either commercially or non-commercially insured patients (Table 3), but patients with non-commercial insurance were hospitalized at markedly higher rates (p < 0.0001) in both years.
Table 3
Factors influencing hemoglobin A1c, linear model
|
Estimate
|
Standard
Error
|
P
|
Intercept
|
7.71
|
0.14
|
< 0.0001
|
Age,y
|
0.05
|
0.01
|
< 0.0001
|
Gender
|
|
|
|
Male
|
0.00
|
|
|
Female
|
0.11
|
0.06
|
0.06
|
Year
|
|
|
|
2019
|
0.00
|
|
|
2020
|
0.04
|
0.06
|
NS
|
Insurance
|
|
|
|
Commercial
|
0.00
|
|
|
Non-commercial
|
0.62
|
0.07
|
< 0.0001
|
Race/ethnicity
|
|
|
|
White
|
0.00
|
|
|
Black
|
1.31
|
0.09
|
< 0.0001
|
Hispanic
|
0.41
|
0.08
|
< 0.0001
|
Other
|
-0.10
|
0.12
|
NS
|
CGM use
|
|
|
|
Yes
|
0.00
|
|
|
No
|
0.87
|
0.07
|
< 0.0001
|
Using data from October 2014 to October 2017, we had previously developed a predictive model for hospital admissions incorporating hospitalizations in the prior 12 months, HbA1c and non-commercial insurance as factors (1). To see if the model retained discrimination (i.e., predictive power) under the changed circumstances of the pandemic, we used data from the 2019 period to predict hospitalization in the 2020 period. As assessed by the area under the receiver operator characteristic curve (ROC AUC), discrimination actually improved from 0.746 in the original training dataset (1) to 0.761 in the present study. In the original training dataset, a risk score of 0.3 had 95% specificity and 29% sensitivity to predict hospitalization; in the present study, the same threshold had 94% specificity and 32% sensitivity. Thus, model performance was essentially unchanged during the pandemic.
The effects of the pandemic on glycemic control were examined in a generalized linear model (Table 4). Increasing age, non-commercial insurance, Black and Hispanic race/ethnicity, and non-utilization of continuous glucose monitors (CGM) were all associated with higher HbA1c, but there was no difference between the 2019 and 2020 groups. There was no change in CGM utilization in patients with commercial insurance (61.8% in 2019 and 61.4% in 2020), but CGM utilization by patients with non-commercial insurance increased markedly from 24.5% in 2019 to 35.7% in 2020 (p = 0.001), probably because Texas Medicaid began approving reimbursement for CGM in April 2020. CGM percent time in range was strongly correlated with HbA1c (R2 = 0.49, p < 0.0001). Similar to the findings regarding HbA1c, time in range among patients utilizing CGM was lower in those with non-commercial insurance and in Black and Hispanic patients; it improved slightly from 2019 to 2020.
Table 4
Factors influencing CGM time in range (%), linear model
|
Estimate
|
Standard
Error
|
P
|
Intercept
|
45.59
|
2.02
|
< 0.0001
|
Age,y
|
-0.19
|
0.12
|
NS
|
Gender
|
|
|
|
Male
|
0.00
|
|
|
Female
|
-0.03
|
0.92
|
NS
|
Year
|
|
|
|
2019
|
0.00
|
|
|
2020
|
1.93
|
0.92
|
0.04
|
Insurance
|
|
|
|
Commercial
|
0.00
|
|
|
Non-commercial
|
-7.50
|
1.15
|
< 0.0001
|
Race/ethnicity
|
|
|
|
White
|
0.00
|
|
|
Black
|
-7.60
|
1.49
|
< 0.0001
|
Hispanic
|
-2.96
|
1.40
|
0.03
|
Other
|
1.74
|
1.83
|
NS
|
We routinely screen for depression in our patients 10 years of age and older using the Patient Health Questionaire-9 (PHQ9); the proportion of screened patients in the entire clinic population decreased in 2020 from 58.5–41.5% (p < 0.0001) because we did not attempt to have patients complete the questionnaire online. Among those screened, the only demographic factor associated with increased scores was female gender; there was no significant change from 2019 to 2020 (Table 5).
Table 5
Factors influencing PHQ9, linear model
|
Estimate
|
Standard
Error
|
P
|
Intercept
|
2.03
|
0.57
|
0.0003
|
Age
|
0.00
|
0.03
|
NS
|
Year
|
|
|
|
2019
|
0.00
|
|
|
2020
|
-0.20
|
0.17
|
NS
|
Insurance
|
|
|
|
Commercial
|
|
|
|
Non-commercial
|
0.19
|
0.19
|
NS
|
Gender
|
|
|
|
Male
|
0.00
|
|
|
Female
|
0.74
|
0.17
|
< 0.0001
|
Race/ethnicity
|
|
|
|
White
|
0.00
|
|
|
Black
|
0.20
|
0.24
|
NS
|
Hispanic
|
0.13
|
0.22
|
NS
|
Other
|
-0.20
|
0.36
|
NS
|