The MEDIC data repository identified 4355 patient visits to the PED between November 1, 2017 and April 30, 2018. A total of 181 visits were excluded for the following reasons: presented to a satellite suburban PED, 44; resided out-of-state, 12; no matching MCIR record, 125. Among 4174 eligible patient visits, 3392 unique patients were identified and 1049 / 3392 (30.9%) received an influenza vaccine during the study period. Overall, the majority of PED visits were by African Americans (80.7%), males (57.8%), children under 8 years of age (54.8%), and children with public or no insurance (83.0%). The median patient age for vaccinated and unvaccinated groups was 7 years (IQR 4.0-11.0). Overall, 23.9% of PED visits were associated with influenza testing. Of those with influenza testing, a positive influenza test result was documented at the time of the visit in 22.9% of encounters, with 83.4% of positive results due to influenza A. Additional demographic and clinical characteristics by vaccination status for the PED visit are reported in Table 1.
Table 1
Study population demographics and clinical characteristics according to vaccination status.
| Overall n = 4174 n (%) | Vaccinated n = 1323 n (%) | Unvaccinated n = 2851 n (%) | p-value |
Gender Female Male | 1761 (42.2) 2413 (57.8) | 548 (41.4) 775 (58.6) | 1213 (42.5) 1638 (57.5) | 0.493 |
Age, years 2-4 5-11 12-18 | 1239 (29.7) 2008 (48.1) 927 (22.2) | 425 (32.1) 602 (45.5) 296 (22.4) | 814 (28.6) 1406 (49.3) 631 (22.1) | 0.036 |
Race American Indian/Alaska Native Asian Black/African American White Patient declined or unknown | 3 (0.1) 12 (0.3) 3369 (80.7) 273 (6.5) 517 (12.4) | 3 (0.2) 6 (0.5) 970 (73.3) 109 (8.2) 235 (17.8) | 0 6 (0.2) 2399 (84.1) 164 (5.8) 282 (9.9) | <0.001 |
Ethnicity Hispanic/Latino Not Hispanic/Latino | 179 (4.3) 3995 (95.7) | 81 (6.1) 1242 (93.9) | 98 (3.4) 2753 (96.6) | <0.001 |
Detroit Neighborhood (n=3069) East Riverfront Northeast Southwest Westside Woodward Corridor | 294 (9.6) 803 (26.2) 319 (10.4) 1222 (39.8) 431 (14.0) | 98 (10.5) 251 (27.0) 122 (13.1) 335 (36.0) 124 (13.3) | 196 (9.2) 552 (25.8) 197 (9.2) 887 (41.5) 307 (14.4) | 0.002 |
Payer Type Private Public No insurance | 711 (17.0) 3306 (79.2) 157 (3.8) | 247 (18.7) 1047 (79.1) 29 (2.2) | 464 (16.3) 2259 (79.2) 128 (4.5) | <0.001 |
PED Disposition Admission* Discharge Other | 1066 (25.5) 3067 (73.5) 41 (1.0) | 364 (27.5) 943 (71.3) 16 (1.2) | 702 (24.6) 2124 (74.5) 25 (0.9) | 0.072 |
Presenting for Asthma Symptoms Yes No | 1753 (42.0) 2421 (58.0) | 532 (40.2) 791 (59.8) | 1221 (42.8) 1630 (57.2) | 0.111 |
Influenza Test Documented Yes No | 999 (23.9) 3175 (76.1) | 333 (25.2) 990 (74.8) | 666 (23.4) 2185 (76.6) | 0.202 |
Influenza Test Result (n=999) Positive Negative Unknown | 229 (22.9) 767 (76.8) 3 (0.3) | 65 (19.5) 266 (79.9) 2 (0.6) | 164 (24.6) 501 (75.2) 1 (0.2) | 0.098 |
Influenza Type (n=229) Influenza A positive Influenza B positive | 191 (83.4) 38 (16.6) | 56 (86.2) 9 (13.8) | 135 (82.3) 29 (17.7) | 0.482 |
*Includes observation and inpatient hospital admissions. |
The majority of PED visits were from patients residing in the Southeast region of the state (4161, 99.7%), with 11 visits (0.3%) from patients in the Central region, and 2 visits (<0.1%) from patients in the Western region. Most patients residing in the Southeast region of the state were living in the city of Detroit (2482 / 3392 patients, 73.2%). Within the city of Detroit, the relative majority of patients resided within the Westside neighborhood (40.0%), followed by Northeast (26.5%), Woodward Corridor (14.1%), Southwest (10.7%), and East Riverfront (8.7%). As shown in Table 1, vaccination status varied by neighborhood: the greatest proportion of unvaccinated visits were by patients who lived in the Westside (41.5%). Within each neighborhood, the highest proportion of vaccinated patients lived in the Southwest (100 / 266, 37.6%), followed by East Riverfront (66 / 216, 30.6%), Northeast (201 / 658. 30.5%), Woodward Corridor (100 / 349, 28.7%), and Westside (270 / 993, 27.2%).
Among vaccinated patients (Table 2, n=1049), there were no notable differences in the timing of the vaccination with 49.1% receiving an “early” vaccine and 50.9% receiving a “late” vaccine. Most vaccines were administered in a primary care setting (84.3%) and through publicly funded campaigns (76.8%). The majority of vaccinated patients had received the influenza vaccine prior to the PED visit (n = 1047, 79.1%).
Table 2
Vaccine administration characteristics for vaccinated children, n (%) (n=1049).
Vaccine administration date* Early Late | 515 (49.1) 534 (50.9) |
Location administering vaccine Primary Care Subspecialty or Other Clinic± Hospital Local Health Department Pharmacy | 884 (84.3) 81 (7.7) 42 (4.0) 23 (2.2) 19 (1.8) |
Vaccine administration publicly funded Yes No | 806 (76.8) 243 (23.2) |
*Early vaccine administration date (September-October) and late vaccine administration date (November-April) determined according to CDC recommendations [24]. |
±Includes vaccines administered at the following locations: Correctional clinic, Local health department satellite clinic, School-based health center, Specialty clinic, Teen health center. |
Although many visits captured patients with a history of asthma (58.0%), a greater proportion of unvaccinated patients were presenting to the PED for asthma-related symptoms (42.8% vs. 40.2%) and tested positive for influenza (24.6% vs. 19.5%). Among patients presenting to the PED for asthma related symptoms (n = 1753, 42.0%), only 30.3% (p-value 0.111) had received an influenza vaccination for the 2017-2018 season and 75.8% (p-value 0.013) of those vaccinated were vaccinated prior to the PED visit (data not shown). When stratified by age (Table 3), the proportion of PED visits for an acute asthma exacerbation decreased with age, from 55.5% of visits among children aged 2-4 years to 24.9% of visits among children 12-18 years. Notably, a smaller proportion of visits for ages 5-11 (30.0%) were by children who had received the influenza vaccine. Over half of visits across all age groups were by children with late vaccine administration.
Table 3
Asthma and influenza characteristics by patient age groups.
| Age in years | p-value |
2-4 | 5-11 | 12-18 |
Presenting for asthma exacerbation Yes No | 688 (55.5) 551 (44.5) | 834 (41.5) 1174 (58.5) | 231 (24.9) 696 (75.1) | <0.001 |
Influenza case Yes No | 80 (6.5) 1159 (93.5) | 127 (6.3) 1881 (93.7) | 28 (3.0) 899 (97.0) | <0.001 |
Influenza test documented Yes No | 406 (32.8) 833 (67.2) | 453 (22.6) 1555 (77.4) | 140 (15.1) 787 (84.9) | <0.001 |
Influenza test result (n=999) Positive Negative Unknown | 75 (18.5) 329 (81.0) 2 (0.5) | 126 (27.8) 326 (72.0) 1 (0.2) | 28 (20.0) 112 (80.0) 0 | 0.015 |
Vaccinated Yes No | 425 (34.3) 814 (65.7) | 602 (30.0) 1406 (70.0) | 296 (31.9) 631 (68.1) | 0.036 |
Vaccinated before PED visit (n=1323) Yes No | 342 (80.5) 83 (19.5) | 464 (77.1) 138 (22.9) | 241 (81.4) 55 (18.6) | 0.230 |
Vaccine administration date* (n=1049) Early Late | 166 (49.6) 169 (50.4) | 229 (48.9) 239 (51.1) | 120 (48.8) 126 (51.2) | 0.979 |
Reported by overall patient visits to the pediatric emergency department (PED) (n=4174) unless otherwise noted. |
*Early vaccine administration date (September-October) and late vaccine administration date (November-April) determined according to CDC recommendations [24]. |
Additional analyses were performed by gender, race, and primary payer type. None of the analyses by gender were statistically significant. Conversely, whether the patient presented to the PED for asthma related symptoms (p-value <0.001), was diagnosed with influenza (p-value 0.023), received an influenza vaccination (p-value <0.001), or received an “early” influenza vaccine (p-value 0.008) were all statistically significant by racial group. Among PED visits for patients of African American race, a greater proportion presented to the PED for asthma-related symptoms (44.3%), were diagnosed with influenza (5.3%), were unvaccinated (71.2%), and received a “late” influenza vaccination (54.5%) than non-Hispanic white patients. Although not statistically significant (p-value 0.053), a smaller proportion of PED visits for African American patients were associated with influenza testing (23.2%). Finally, by primary payer type, both influenza testing documented at the time of the PED visit (p-value 0.021) and vaccination status (<0.001) were statistically significant. Although only a minority of PED visits were associated with influenza testing, 14.6% of those with no insurance had influenza testing documented in contrast to 24.6% of those with private insurance and 24.2% of those with public insurance. A greater proportion of visits with private insurance were associated with vaccination (34.7%) in contrast to visits with public (31.7%) or no insurance (18.5%).
Spatial analyses at the Census Tract level found high levels of clustering. Figure 1 displays the raw rate of influenza vaccination coverage by percentile by Census Tract in Detroit. Vaccination coverage among patients across Detroit ranged from a low of 10–100% in Census Tracts across the city. Vaccination coverage of <50% is particularly notable in the areas denoted in blue on Figure 1. There were significant degrees of clustering among vaccinated (Figure 2) and unvaccinated (Figure 3) patients with unvaccinated patients generating the greatest number of hotspots in the city.
The Global Moran’s I statistic (Table 4) identified two significant findings for spatial autocorrelation among unvaccinated patients: Black or African American race and being in the age 5 to 11 group. These two significant findings were also examined using LISA to pinpoint hotspots (Figure 4 and Figure 5).
Table 4
Results of statistical analyses of Global Moran’s I measuring spatial autocorrelation
| Vaccinated | Unvaccinated |
| Moran’s I | p-value* | Moran’s I | p-value* |
Flu Case | -0.015 | 0.306 | -0.002 | 0.496 |
Asthma Exacerbation | 0.024 | 0.178 | 0.052 | 0.046 |
White | 0.048 | 0.041 | -0.034 | 0.078 |
Black | -0.029 | 0.169 | 0.080 | 0.014 |
Ages 2 to 4 | -0.009 | 0.405 | 0.028 | 0.151 |
Ages 5 to 11 | 0.039 | 0.094 | 0.070 | 0.017 |
Ages 12 to 18 | 0.019 | 0.219 | 0.031 | 0.124 |
*Indicates a pseudo p-value at permutation of 999 |