Results

We identified 10,450 weighted adult hospitalizations with NHL, MM, or ALL who received CAR T-cell therapy between 2017 and 2022. Of these, 7620 (72.9%) had NHL, 2075 (19.9%) had MM, and 755 (7.2%) had ALL. DIC was diagnosed in 220 hospitalizations (2.1%), while 10,230 (97.9%) did not have DIC. Among patients who developed DIC, 135 (61.4%) had NHL, 50 (22.7%) had MM, and 35 (15.9%) had ALL. Compared with those without DIC, patients with DIC had a significantly higher proportion in the lowest income quartile (29.6% vs 17.3%; P = .038), self-pay status (6.8% vs 0.8%, P < .001), and sepsis (4.6% vs 0.8%; P = .007), along with a lower proportion of Hispanic patients (2.3% vs 12.2%; P = .046) (Table 1).

Characteristic | Patients with DIC | Patients without DIC | P
Patient characteristics
No. (%) of patients | 220 (2.1) | 10,230 (97.9) | —
Mean age (y) | 57.6 | 60.3 | .209
Female | 75 (34.1) | 3920 (38.3) | .615
Race/ethnicity
White | 165 (75.0) | 7,415 (72.5) | .722
Black | 20 (9.1) | 725 (7.1) | .618
Hispanic | 5 (2.3) | 1250 (12.2) | .046
Asian or Pacific Islander | 15 (6.8) | 415 (4.1) | .460
Native American | 5 (2.3) | 45 (0.4) | .093
Other | 10 (4.6) | 380 (3.7) | .775
Charlson Comorbidity index score
2 | 25 (11.4) | 1260 (12.3) | .839
≥3 | 195 (88.6) | 8970 (87.7) | .839
Patient’s zip code income quartile
First | 65 (29.6) | 1765 (17.3) | .038
Second | 60 (27.3) | 2200 (21.5) | .384
Third | 40 (18.2) | 2740 (26.8) | .195
Fourth | 55 (25.0) | 3525 (34.5) | .178
Insurance type
Medicare | 70 (31.8) | 4265 (41.7) | .142
Medicaid | 20 (9.1) | 750 (7.3) | .653
Private | 115 (52.3) | 5130 (50.2) | .761
Self-pay | 15 (6.8) | 85 (0.8) | <.001
Hospital teaching status/location
Rural | 0 (0.0) | 20 (0.2) | .830
Urban nonteaching | 0 (0.0) | 35 (0.3) | .876
Urban teaching | 220 (100.0) | 10,175 (99.5) | .815
Comorbidities
Sepsis | 10 (4.6) | 80 (0.8) | .007
Cytokine release syndrome | 90 (40.9) | 4130 (40.4) | .945
SARS-CoV-2 | 5 (2.3) | 130 (1.3) | .516
Dyslipidemia | 55 (25.0) | 2565 (25.1) | .992
Hypertension | 80 (36.4) | 4495 (43.9) | .337
Atrial fibrillation | 35 (15.9) | 1220 (11.9) | .432
Long-term anticoagulation use | 10 (4.6) | 1050 (10.3) | .220
Long-term antiplatelet use | 10 (4.6) | 731 (7.1) | .516

In-hospital mortality based on DIC status

Among patients who developed DIC, the in-hospital mortality rate was 22.7%, markedly higher than patients who did not develop DIC, which was 2.4% (Table 2). In both univariable and multivariable analyses, adjusting for patient- and hospital-level confounders, DIC was independently associated with increased in-hospital mortality (adjusted odds ratio [aOR], 10.96; 95% confidence interval [CI], 4.78-25.09; P < .001) (Table 3). These findings remained robust in a secondary model adjusted only for variables associated with increased mortality at a univariable significance threshold of P < .20, including race (Asian or Pacific Islander), CCI, dyslipidemia, atrial fibrillation, and long-term anticoagulation use (aOR, 10.74; 95% CI, 4.88-23.64; P < .001) (¹).

Outcome | Patients with DIC | Patients without DIC
Primary outcome
Died | 50 (22.7) | 245 (2.4)
Secondary outcomes
AKI | 105 (47.7) | 1410 (13.8)
RF | 85 (38.6) | 600 (5.9)
Shock | 95 (43.2) | 375 (3.7)
RRT | 25 (11.4) | 135 (1.3)
MV | 90 (40.9) | 315 (3.1)
Vasopressor | 50 (22.7) | 220 (2.2)
VTE | 15 (6.8) | 255 (2.5)
ACS | 5 (2.3) | 50 (0.5)
GIH | 35 (15.9) | 135 (1.3)
ICH | 10 (4.6) | 120 (1.2)
RBC | 50 (22.7) | 980 (9.6)
Platelet | 45 (20.5) | 460 (4.5)
Cryoprecipitate/FFP | 40 (18.2) | 190 (1.9)

Outcome | Crude OR (95% CI) | Adjusted OR (95% CI)
Died | 11.99 (5.91-24.30) | 10.96 (4.78-25.09)
AKI | 5.71 (3.12-10.44) | 6.01 (3.05-11.82)
RF | 10.11 (5.24-19.48) | 9.96 (4.90-20.26)
Shock | 19.97 (10.00-39.89) | 18.55 (8.60-40.01)
RRT | 9.59 (3.68-24.99) | 10.63 (3.30-34.23)
MV | 21.79 (11.28-42.10) | 22.94 (11.21-46.93)
Vasopressor | 13.38 (6.13-29.19) | 13.52 (5.55-32.92)
VTE | 2.86 (0.87-9.42) | 2.94 (0.82-10.52)
ACS | 4.73 (0.53-42.16) | 4.39 (0.24-82.00)
GIH | 14.15 (5.82-34.39) | 12.41 (4.36-35.35)
ICH | 4.01 (0.93-17.33) | 3.88 (0.74-20.20)
RBC | 2.78 (1.28-6.02) | 2.86 (1.30-6.29)
Platelet | 5.46 (2.39-12.46) | 5.67 (2.32-13.86)
Cryoprecipitate/FFP | 11.74 (4.65-29.66) | 12.70 (4.48-35.98)

Organ dysfunction and critical care support

Among patients who developed DIC, the rates of AKI (47.7% vs 13.8%), RF (38.6% vs 5.9%), and shock (43.2% vs 3.7%) were significantly higher (Table 2). In both univariable and multivariable analyses, adjusting for patient- and hospital-level confounders, DIC was independently associated with the development of AKI (aOR, 6.01; 95% CI, 3.05-11.82; P < .001), RF (aOR, 9.96; 95% CI, 4.90-20.26; P < .001), and shock (aOR, 18.55; 95% CI, 8.60-40.01; P < .001) (Table 3).

Furthermore, patients with DIC demonstrated significantly higher rates of critical care interventions, including RRT (11.4% vs 1.3%), MV (40.9% vs 3.1%), and vasopressor use (22.7% vs 2.2%) (Table 2). In both univariable and multivariable models, DIC remained independently associated with increased odds of RRT (aOR, 10.63; 95% CI, 3.30-34.23; P < .001), MV (aOR, 22.94; 95% CI, 11.21-46.93; P < .001), and vasopressor (aOR, 13.52; 95% CI, 5.55-32.92; P < .001) utilization (Table 3).

Thrombosis, hemorrhage, and blood product utilization

Patients with DIC experienced a higher rate of thromboembolic complications than their counterparts without DIC. Specifically, the rate of VTE was elevated (6.8% vs 2.5%), as was the rate of ACS (2.3% vs 0.5%) (Table 2). However, after adjusting for patient- and hospital-level confounders, DIC was not significantly associated with either VTE (aOR, 2.94; 95% CI, 0.82-10.52; P = .097) or ACS (aOR, 4.39; 95% CI, 0.24-82.0; P = .321) (Table 3).

In addition, patients with DIC exhibited significantly higher rates of GIH (15.9% vs 1.3%) and ICH (4.6% vs 1.2%) than those without DIC (Table 2). However, after adjusting for potential confounders, DIC only remained significantly associated with increased odds of GIH (aOR, 12.41; 95% CI, 4.36-35.35; P < .001). The association between DIC and ICH did not reach statistical significance (aOR, 3.88; 95% CI, 0.74-20.20; P = .108) (Table 3).

DIC was also associated with significantly higher rates of blood product transfusion. Specifically, transfusion rates were higher for RBC (22.7% vs 9.6%), platelets (20.5% vs 4.5%), and cryoprecipitate/FFP (18.2% vs 1.9%) transfusions (Table 2). After adjusting for patient- and hospital-level confounders, DIC remained independently associated with increased need for RBC (aOR, 2.86; 95% CI, 1.30-6.29; P = .009), platelet (aOR, 5.67; 95% CI, 2.32-13.86; P < .001), and cryoprecipitate/FFP (aOR, 12.70; 95% CI, 4.48-35.98; P < .001) transfusions (Table 3).

Hospital length of stay and total hospitalization charges based on DIC status

The crude mean LOS and TOTCHG for CAR T-cell therapy recipients with and without DIC are presented in Table 4. After adjustment for patient- and hospital-level confounders, DIC was independently associated with increased LOS (β = +17.6 days; 95% CI, 10.16-25.08; P < .001) and higher TOTCHG (β = +$523,323; 95% CI, $82,053-$964,593; P = .020).

Outcome | Patients with DIC | Patients without DIC
Crude mean LOS (d) | 36.0 (28.6–43.3) | 17.1 (16.4–17.9)
Crude mean TOTCHG ($) | 1,747,822 (1,302,865-2,192,780) | 1,235,012 (1,090,737-1,379,287)

Footnotes

1. The online version contains supplementary material available at https://doi.org/10.1016/j.rpth.2026.106886. ↩