Selected baseline characteristics of IVIg users and nonusers after propensity matching
| Covariate . | IVIg users* (n = 2724) . | Nonusers* (n = 8035) . | Standardized difference† . |
|---|---|---|---|
| Cancer diagnosis and demographics | |||
| CLL diagnosis‡ | 76% | 76% | 0.00 |
| MM diagnosis‡ | 24% | 24% | 0.00 |
| Years since diagnosis with CLL or MM | 4.44 (3.72) | 4.32 (3.83) | 0.03 |
| Index date calendar year‡ | 0.00 | ||
| 1992-1998 | 18% | 18% | |
| 1999-2004 | 31% | 31% | |
| 2005-2010 | 51% | 51% | |
| Age in years | 75.76 (6.42) | 75.87 (6.55) | −0.02 |
| Female sex | 41% | 42% | −0.00 |
| Health care utilization | |||
| Hospital stay | 64% | 68% | −0.07 |
| Emergency department visit | 57% | 59% | −0.05 |
| Nursing home stay | 5% | 6% | −0.02 |
| Medicare Part D‡ | 18% | 18% | −0.00 |
| Cardiovascular risk factors | |||
| Cerebrovascular disease | 16% | 16% | −0.00 |
| Heart failure | 27% | 29% | −0.03 |
| Coronary heart disease | 36% | 37% | −0.03 |
| Diabetes, complicated | 10% | 10% | −0.01 |
| Diabetes, uncomplicated | 18% | 18% | 0.01 |
| Hypertension, complicated | 14% | 14% | −0.01 |
| Hypertension, uncomplicated | 48% | 47% | 0.00 |
| Peripheral vascular disease | 18% | 19% | −0.02 |
| Chronic kidney disease | 16% | 17% | −0.02 |
| Tobacco use | 11% | 11% | −0.00 |
| Valvular heart disease | 25% | 26% | −0.01 |
| Venous thromboembolism risk factors | |||
| Coagulopathy, congenital | 1% | 1% | −0.01 |
| Coagulopathy, other | 8% | 7% | 0.01 |
| Thrombocytopenia | 30% | 28% | 0.04 |
| Central venous catheter | 9% | 8% | 0.01 |
| Functional limitation | 22% | 22% | −0.01 |
| Other condition causing immobility | 21% | 23% | −0.03 |
| Malnutrition or unintentional weight loss | 14% | 14% | −0.01 |
| Hip or knee surgery | 1% | 1% | −0.02 |
| Other surgery | 6% | 5% | 0.01 |
| VTE (outcome definition) | 7% | 7% | 0.00 |
| VTE (non-outcome definition) | 9% | 9% | 0.00 |
| Blood transfusion | 19% | 17% | 0.05 |
| Other VTE risk factor | 4% | 4% | −0.01 |
| Solid tumor, high VTE risk | 2% | 2% | −0.01 |
| Solid tumor, low VTE risk | 10% | 9% | 0.01 |
| Anticoagulant medication use | 9% | 9% | 0.00 |
| Indicators of infection risk and/or cancer progression | |||
| Systemic corticosteroid use | 42% | 42% | 0.01 |
| Antineoplastic chemotherapy (any) | 59% | 59% | −0.00 |
| Antineoplastic radiation therapy | 8% | 8% | −0.01 |
| Bone marrow failure/aplastic anemia | 28% | 26% | 0.05 |
| Neutropenia | 27% | 26% | 0.02 |
| Influenza | 4% | 4% | 0.01 |
| Pneumonia (outpatient diagnosis only) | 18% | 19% | −0.04 |
| Pneumonia (inpatient diagnosis) | 32% | 32% | −0.01 |
| Chronic bronchitis | 18% | 18% | −0.00 |
| Sepsis or septic shock | 21% | 20% | 0.02 |
| Urinary tract infection | 26% | 26% | 0.00 |
| Other infection | 9% | 9% | −0.02 |
| Covariate . | IVIg users* (n = 2724) . | Nonusers* (n = 8035) . | Standardized difference† . |
|---|---|---|---|
| Cancer diagnosis and demographics | |||
| CLL diagnosis‡ | 76% | 76% | 0.00 |
| MM diagnosis‡ | 24% | 24% | 0.00 |
| Years since diagnosis with CLL or MM | 4.44 (3.72) | 4.32 (3.83) | 0.03 |
| Index date calendar year‡ | 0.00 | ||
| 1992-1998 | 18% | 18% | |
| 1999-2004 | 31% | 31% | |
| 2005-2010 | 51% | 51% | |
| Age in years | 75.76 (6.42) | 75.87 (6.55) | −0.02 |
| Female sex | 41% | 42% | −0.00 |
| Health care utilization | |||
| Hospital stay | 64% | 68% | −0.07 |
| Emergency department visit | 57% | 59% | −0.05 |
| Nursing home stay | 5% | 6% | −0.02 |
| Medicare Part D‡ | 18% | 18% | −0.00 |
| Cardiovascular risk factors | |||
| Cerebrovascular disease | 16% | 16% | −0.00 |
| Heart failure | 27% | 29% | −0.03 |
| Coronary heart disease | 36% | 37% | −0.03 |
| Diabetes, complicated | 10% | 10% | −0.01 |
| Diabetes, uncomplicated | 18% | 18% | 0.01 |
| Hypertension, complicated | 14% | 14% | −0.01 |
| Hypertension, uncomplicated | 48% | 47% | 0.00 |
| Peripheral vascular disease | 18% | 19% | −0.02 |
| Chronic kidney disease | 16% | 17% | −0.02 |
| Tobacco use | 11% | 11% | −0.00 |
| Valvular heart disease | 25% | 26% | −0.01 |
| Venous thromboembolism risk factors | |||
| Coagulopathy, congenital | 1% | 1% | −0.01 |
| Coagulopathy, other | 8% | 7% | 0.01 |
| Thrombocytopenia | 30% | 28% | 0.04 |
| Central venous catheter | 9% | 8% | 0.01 |
| Functional limitation | 22% | 22% | −0.01 |
| Other condition causing immobility | 21% | 23% | −0.03 |
| Malnutrition or unintentional weight loss | 14% | 14% | −0.01 |
| Hip or knee surgery | 1% | 1% | −0.02 |
| Other surgery | 6% | 5% | 0.01 |
| VTE (outcome definition) | 7% | 7% | 0.00 |
| VTE (non-outcome definition) | 9% | 9% | 0.00 |
| Blood transfusion | 19% | 17% | 0.05 |
| Other VTE risk factor | 4% | 4% | −0.01 |
| Solid tumor, high VTE risk | 2% | 2% | −0.01 |
| Solid tumor, low VTE risk | 10% | 9% | 0.01 |
| Anticoagulant medication use | 9% | 9% | 0.00 |
| Indicators of infection risk and/or cancer progression | |||
| Systemic corticosteroid use | 42% | 42% | 0.01 |
| Antineoplastic chemotherapy (any) | 59% | 59% | −0.00 |
| Antineoplastic radiation therapy | 8% | 8% | −0.01 |
| Bone marrow failure/aplastic anemia | 28% | 26% | 0.05 |
| Neutropenia | 27% | 26% | 0.02 |
| Influenza | 4% | 4% | 0.01 |
| Pneumonia (outpatient diagnosis only) | 18% | 19% | −0.04 |
| Pneumonia (inpatient diagnosis) | 32% | 32% | −0.01 |
| Chronic bronchitis | 18% | 18% | −0.00 |
| Sepsis or septic shock | 21% | 20% | 0.02 |
| Urinary tract infection | 26% | 26% | 0.00 |
| Other infection | 9% | 9% | −0.02 |
For categorical variables, the percentage of patients with a given condition is shown. For continuous variables, the mean (standard deviation) is shown. To account for the use of variable-ratio matching of k users to each IVIg user, each IVIg user received a weight of 1 and each nonuser a weight of 1/k. Distributions for all variables included as propensity score components may be found in supplemental Table 4. VTE, venous thromboembolic event.
Standardized differences represent the mean between-group difference divided by the pooled within-group standard deviation, so that a value of 1.0 represents a 1 − SD difference.
The matching process required exact matching of nonusers to IVIg users on cancer type (CLL or MM), the index date (i.e., the IVIg user’s treatment initiation date), and Part D utilization status during the 1-year lookback period.