Abstract
Clostridium difficile infections (CDI) are the leading cause of infectious diarrhea among pts undergoing stem cell transplantation (SCT). Data from large pt samples evaluating the trends, risk factors, hospitalization outcomes are lacking. We have analyzed the Nationwide Inpt Sample (NIS) database to evaluate the trends in the incidence and the impact of CDI on hospitalization outcomes among SCT recipients.
We used the NIS database to obtain the hospitalization data for primary procedures of autologous and allogeneic SCT between 01/2001 until 12/2010. We performed separate multivariate logistic regression analyses to evaluate risk factors of CDI in SCT pts and hospitalization outcomes including length of stay (LOS), in-hospital mortality (IHM) and hospitalization charges. Comorbidites were identified using software that created measures reported by Elixhauser et al.
The incidence of CDI among transplant pts has remained relatively stable over time (Fig 1), with a higher cumulative incidence of CDI among recipients of allogeneic grafts (allogeneic vs. autologous SCT: 8.29% vs. 5.63%; p<0.001). There was an association between CDI and the presence of multiple comorbidities among autologous SCT pts (p=0.05), and a similar trend was seen in allogeneic transplant pts. Mean LOS (CDI vs. none: autologous 26 vs. 20 days, p<0.0001; allogeneic 42 vs 30 days, p<0.001; respectively), IHM (CDI vs. none: autologous, 4.0 vs. 2.2%, p=0.0003; allogeneic 12.7 vs 9.2%, p=0.0008) and mean hospitalization charges (CDI vs. none: autologous $208,025 vs. $145,250, p<0.0001; allogeneic $367,415 vs $270,366 p<0.0001; respectively) were significantly higher among SCT pts with CDI (Table 1).
Trend of CDI in SCT recipients
Pt characteristics, multivariate regression models for evaluating risk and outcomes of SCT pts with CDI
| Autologous SCT | Allogeneic SCT | |||||
|---|---|---|---|---|---|---|
| CDI (%) (N=3772) | SCT patients (%) (N=63202) | p-value | CDI (%) (N=3440) | SCT patients (N=38066) | p-value | |
| Mean age (SE) | 54.3 (0.52) | 52.02 (0.27) | <0.0001 | 45.07 (0.64) | 46.57 (0.38) | <0.0098 |
| Male | 2311 (61.3) | 37038 (58.6) | 0.1596 | 2188 (63.6) | 21906 (57.6) | 0.0005 |
| White | 2369 (77.4) | 26124 (74.3) | 0.1089 | 2199 (77.7) | 23695 (78) | 0.847 |
| Comorbidities ≥1 | 2945 (78) | 46872 (74.2) | 0.05 | 2607 (75.8) | 27712 (72.8) | 0.1073 |
| Medicare | 945 (25.1) | 12460 (19.8) | <0.0001 | 310 (9.0) | 4111 (10.8) | 0.183 |
| Medicaid | 330 (8.8) | 6818 (10.8) | 0.1161 | 474 (13.8) | 3996 (10.5) | 0.0126 |
| Private Insurance | 2320 (61.6) | 40209 (63.8) | 0.2483 | 2486 (72.4) | 27601 (72.7) | 0.8798 |
| Mean LOS (SE)* | 26 (0.71) | 20 (0.27) | <0.0001* | 42 (1.59) | 30 (0.68) | <0.0001* |
| IHM | 152 (4.0) | 1330 (2.1) | 0.0003 | 438 (12.7) | 3360 (8.8) | 0.0008 |
| Charge Mean (SE) | 208025 (8370.6) | 145250 (4311.3) | <0.0001* | $367415 (15033) | $270366 (11835) | <0.0001* |
| Multivariate logistic regression model evaluating risk of CDI in SCT patients | ||||||
| Variable | Odds Ratio | Ref | p-value | Odds Ratio | Ref | p-value |
| ≥65 | 1.4 (0.987-1.985) | Age <40 (Ref) | 0.1945 | 0.637 (0.426-0.953) | 0.0834 | |
| Male | 1.25 (1.054-1.483) | Female (Ref) | 0.0103 | 1.351 (1.157-1.576) | Female (Ref) | 0.0001 |
| Medicare | 1.149 (0.878-1.502) | Private (Ref) | 0.0439 | 0.85 (0.581-1.244) | Private (Ref) | 0.4605 |
| Medicaid | 0.887 (0.642-1.226) | 0.7673 | 1.269 (0.920-1.749) | 0.0272 | ||
| No Comorbidities | 0.831 (0.650-1.063) | Comorbidities ≥1 | 0.1406 | 0.829 (0.672-1.024) | Comorbidities ≥1 | 0.082 |
| Autologous SCT | Allogeneic SCT | |||||
|---|---|---|---|---|---|---|
| CDI (%) (N=3772) | SCT patients (%) (N=63202) | p-value | CDI (%) (N=3440) | SCT patients (N=38066) | p-value | |
| Mean age (SE) | 54.3 (0.52) | 52.02 (0.27) | <0.0001 | 45.07 (0.64) | 46.57 (0.38) | <0.0098 |
| Male | 2311 (61.3) | 37038 (58.6) | 0.1596 | 2188 (63.6) | 21906 (57.6) | 0.0005 |
| White | 2369 (77.4) | 26124 (74.3) | 0.1089 | 2199 (77.7) | 23695 (78) | 0.847 |
| Comorbidities ≥1 | 2945 (78) | 46872 (74.2) | 0.05 | 2607 (75.8) | 27712 (72.8) | 0.1073 |
| Medicare | 945 (25.1) | 12460 (19.8) | <0.0001 | 310 (9.0) | 4111 (10.8) | 0.183 |
| Medicaid | 330 (8.8) | 6818 (10.8) | 0.1161 | 474 (13.8) | 3996 (10.5) | 0.0126 |
| Private Insurance | 2320 (61.6) | 40209 (63.8) | 0.2483 | 2486 (72.4) | 27601 (72.7) | 0.8798 |
| Mean LOS (SE)* | 26 (0.71) | 20 (0.27) | <0.0001* | 42 (1.59) | 30 (0.68) | <0.0001* |
| IHM | 152 (4.0) | 1330 (2.1) | 0.0003 | 438 (12.7) | 3360 (8.8) | 0.0008 |
| Charge Mean (SE) | 208025 (8370.6) | 145250 (4311.3) | <0.0001* | $367415 (15033) | $270366 (11835) | <0.0001* |
| Multivariate logistic regression model evaluating risk of CDI in SCT patients | ||||||
| Variable | Odds Ratio | Ref | p-value | Odds Ratio | Ref | p-value |
| ≥65 | 1.4 (0.987-1.985) | Age <40 (Ref) | 0.1945 | 0.637 (0.426-0.953) | 0.0834 | |
| Male | 1.25 (1.054-1.483) | Female (Ref) | 0.0103 | 1.351 (1.157-1.576) | Female (Ref) | 0.0001 |
| Medicare | 1.149 (0.878-1.502) | Private (Ref) | 0.0439 | 0.85 (0.581-1.244) | Private (Ref) | 0.4605 |
| Medicaid | 0.887 (0.642-1.226) | 0.7673 | 1.269 (0.920-1.749) | 0.0272 | ||
| No Comorbidities | 0.831 (0.650-1.063) | Comorbidities ≥1 | 0.1406 | 0.829 (0.672-1.024) | Comorbidities ≥1 | 0.082 |
Multivariate logistic regression model evaluating hospitalization outcomes in SCT patients with CDI #
| CDI | Absence | Presence | p-value | CDI | Absence | Presence | p-value |
|---|---|---|---|---|---|---|---|
| LOS >19 days | 0.382 (0.31-0.466) | 1.00 (Ref) | <0.0001 | LOS > 27 days | 0.337 (0.27-0.42) | 1.00 (Ref) | <0.0001 |
| Charge > $125,091 | 0.483 (0.40-0.58) | <0.0001 | Charge > $22,6011 | 0.487 (0.38-0.67) | <0.0001 | ||
| IHM | 0.68 (0.44-1.06) | 0.089 | IHM | 0.661 (0.50-0.87) | 0.0037 |
| CDI | Absence | Presence | p-value | CDI | Absence | Presence | p-value |
|---|---|---|---|---|---|---|---|
| LOS >19 days | 0.382 (0.31-0.466) | 1.00 (Ref) | <0.0001 | LOS > 27 days | 0.337 (0.27-0.42) | 1.00 (Ref) | <0.0001 |
| Charge > $125,091 | 0.483 (0.40-0.58) | <0.0001 | Charge > $22,6011 | 0.487 (0.38-0.67) | <0.0001 | ||
| IHM | 0.68 (0.44-1.06) | 0.089 | IHM | 0.661 (0.50-0.87) | 0.0037 |
SE: Standard error; *adjusted for age, race, sex, comorbidity and payer status.#Median LOS and charges used as cut-off for hospitalization outcomes
The incidence of CDI among both autologous and allogeneic SCTs have remained stable over the last decade, with a slightly higher risk among allogeneic SCT pts. CDI was associated with longer hospital stay, higher charges and higher IHM for both autologous and allogeneic transplant pts. While this analysis does not permit us to directly ascribe differences in outcome to CDI alone, these data suggest that CDI has a major morbid impact in the setting of SCT, and provides a rationale for development of more effective approaches for prevention of CDI in these pt populations.
Kaufman: Onyx: Consultancy; Celgene: Consultancy, Research Funding; Novartis: Consultancy, Research Funding; Janssen: Consultancy; Millenium: Consultancy; Merck: Research Funding. Lonial:Millennium: Consultancy; Celgene: Consultancy; Novartis: Consultancy; BMS: Consultancy; Sanofi: Consultancy; Onyx: Consultancy.
