Table 1.
Baseline characteristics of levetiracetam users versus nonusers before and post propensity score matching
| Characteristica | Before Matching | Post Matching | ||||
|---|---|---|---|---|---|---|
| Levetiracetam Users | Levetiracetam Nonusers | Standardized Differenceb (%) | Levetiracetam Users | Levetiracetam Nonusers | Standardized Differenceb (%) | |
| Total patients, N | 4980 | 1,413,139 | 3980 | 7960 | ||
| Demographics | ||||||
| Mean age±SD, yr | 54±21 | 69±16 | 76 | 55±21 | 55±21 | 0 |
| ≤65 yr | 3092 (62) | 395,028 (28) | 73 | 2403 (60) | 4878 (61) | 2 |
| Women | 2549 (51) | 837,046 (59) | 16 | 2048 (52) | 4013 (50) | 2 |
| Long-term care | 352 (7) | 52,086 (4) | 15 | 273 (7) | 522 (7) | 1 |
| Rural residence | 496 (10) | 167,640 (12) | 6 | 409 (10) | 829 (10) | 0 |
| Income quintilec | ||||||
| 1 (lowest) | 1149 (23) | 335,883 (24) | 2 | 958 (24) | 2009 (25) | 3 |
| 3 (middle) | 1002 (20) | 267,725 (19) | 3 | 808 (20) | 1584 (20) | 1 |
| 5 (highest) | 897 (18) | 246,083 (17) | 2 | 696 (18) | 1375 (17) | 1 |
| Comorbiditiesd | ||||||
| Johns Hopkins ACG System Aggregated Diagnosis Groups, mean±SDe | 10.0±4.3 | 8.2±3.8 | 45 | 9.7±4.3 | 9.8±4.2 | 4 |
| AKI | 16 (0.3) | 2517 (0.2) | 2 | 13 (0.3) | 21 (0.3) | 1 |
| Anxiety disorder and depression | 881 (18) | 186,741 (13) | 12 | 699 (18) | 1495 (19) | 4 |
| Benign brain tumor | 11 (0.2) | 219 (0.0) | 6 | 6 (0.2) | 22 (0.3) | 3 |
| Bipolar disorder | 344 (7) | 76,931 (5) | 6 | 280 (7) | 624 (8) | 4 |
| Brain injury | 220 (4) | 7274 (0.5) | 25 | 153 (4) | 287 (4) | 2 |
| Brain cancer | 329 (7) | 726 (0.1) | 37 | 142 (4) | 167 (2) | 11 |
| Cancerf | 1804 (36) | 441,339 (31) | 11 | 1333 (34) | 2621 (33) | 1 |
| CKD | 549 (11) | 134,314 (10) | 5 | 432 (11) | 830 (10) | 2 |
| Chronic liver disease | 361 (7) | 68,436 (5) | 10 | 273 (7) | 598 (8) | 3 |
| Coronary artery disease, including anginag | 1026 (21) | 382,315 (27) | 15 | 824 (21) | 1668 (21) | 1 |
| Congestive heart failure | 451 (9) | 134,949 (10) | 1 | 362 (9) | 774 (10) | 3 |
| Diabetes | 127 (3) | 49,027 (4) | 5 | 103 (3) | 190 (2) | 2 |
| Epilepsy/seizureh | 3481 (70) | 25,154 (2) | 202 | 2482 (62) | 5019 (63) | 1 |
| Previous 90 d | 1087 (22) | 2772 (0.2) | 74 | 460 (12) | 920 (12) | 0 |
| Previous 90–365 d | 1583 (32) | 6731 (0.5) | 94 | 1091 (27) | 1984 (25) | 6 |
| Hypertension | 1322 (27) | 706,449 (50) | 50 | 1095 (28) | 2148 (27) | 1 |
| Migraine | 618 (12) | 85,403 (6) | 22 | 485 (12) | 987 (12) | 1 |
| Mood disorder | 236 (5) | 27,938 (2) | 15 | 174 (4) | 413 (5) | 5 |
| Multiple sclerosis | 28 (0.6) | 2612 (0.2) | 6 | 15 (0.4) | 92 (1) | 11 |
| Neuropathic pain | 67 (1) | 19,639 (1) | 1 | 47 (1) | 141 (2) | 6 |
| Parkinson disease | 42 (0.8) | 6712 (0.5) | 4 | 36 (0.9) | 136 (2) | 9 |
| Peripheral vascular disease | 54 (1) | 16,050 (1) | 0 | 45 (1) | 109 (1) | 3 |
| Stroke, including TIA | 660 (13) | 36,398 (3) | 40 | 481 (12) | 966 (12) | 0 |
| Trigeminal neuralgia | 80 (2) | 20,422 (1) | 2 | 59 (2) | 172 (2) | 6 |
| Concurrent medication usei | ||||||
| ACE inhibitors or ARBs | 851 (17) | 460,963 (33) | 36 | 706 (18) | 1362 (17) | 2 |
| Antidepressants | 239 (5) | 80,146 (6) | 4 | 199 (5) | 405 (5) | 0 |
| Antiepileptics | 942 (19) | 74,430 (5) | 43 | 703 (18) | 1562 (20) | 6 |
| Antineoplastics | 176 (4) | 37,894 (3) | 5 | 129 (3) | 234 (3) | 2 |
| Antipsychotics | 190 (4) | 49,287 (4) | 2 | 165 (4) | 332 (4) | 0 |
| β blockers | 642 (13) | 304,371 (22) | 23 | 518 (13) | 1046 (13) | 0 |
| Benzodiazepines | 413 (8) | 149,512 (11) | 8 | 309 (8) | 653 (8) | 2 |
| Calcium channel blockers | 556 (11) | 285,732 (20) | 25 | 456 (12) | 899 (11) | 1 |
| Glucocorticoids | 263 (5) | 51,529 (4) | 8 | 176 (4) | 338 (4) | 1 |
| Diuretics—potassium sparing | 58 (1) | 40,865 (3) | 12 | 51 (1) | 98 (1) | 1 |
| Diuretics—thiazides | 225 (5) | 156,167 (11) | 25 | 187 (5) | 318 (4) | 4 |
| Diuretics—loop | 228 (5) | 102,693 (7) | 11 | 185 (5) | 388 (5) | 1 |
| Methotrexate | 20 (0.4) | 10,269 (0.7) | 4 | 15 (0.4) | 44 (0.6) | 3 |
| NSAIDs (excluding ASA) | 141 (3) | 128,727 (9) | 27 | 117 (3) | 219 (3) | 1 |
| Statins | 1052 (21) | 533,226 (38) | 37 | 886 (22) | 1757 (22) | 1 |
| Number of unique medications, mean±SD | 3.4±5.0 | 4.9±4.6 | 31 | 3.4±4.9 | 3.4±5.1 | 0 |
| Antiepileptic use (in prior 120 d)i | ||||||
| Carbamazepine | 162 (3) | 7476 (0.5) | 20 | 125 (3) | 275 (4) | 2 |
| Clobazam | 146 (3) | 1079 (0.1) | 24 | 85 (2) | 159 (2) | 1 |
| Divalproex sodium | 96 (2) | 5104 (0.4) | 15 | 67 (2) | 162 (2) | 3 |
| Gabapentin | 120 (2) | 26,704 (2) | 4 | 89 (2) | 267 (3) | 8 |
| Lacosamide | 56 (1) | 178 (0.0) | 15 | 37 (0.9) | 52 (0.7) | 4 |
| Lamotrigine | 115 (2) | 2248 (0.2) | 20 | 85 (2) | 168 (2) | 0 |
| Phenytoin | 426 (9) | 11,026 (0.8) | 37 | 291 (7) | 677 (9) | 5 |
| Pregabalin | 50 (1) | 21,028 (2) | 4 | 44 (1) | 105 (1) | 2 |
| Primidone | 27 (0.5) | 2215 (0.2) | 7 | 22 (0.6) | 33 (0.4) | 2 |
| Topiramate | 43 (0.9) | 1887 (0.1) | 10 | 31 (0.8) | 62 (0.8) | 0 |
| Valproic acid | 63 (1) | 2104 (0.1) | 13 | 46 (1) | 90 (1) | 0 |
| Antiepileptic use (prescribed on index date) | ||||||
| Carbamazepine | 39 (0.8) | 355 (0.0) | 12 | 23 (0.6) | 34 (0.4) | 3 |
| Clobazam | 43 (0.9) | 37 (0.0) | 13 | 13 (0.3) | 10 (0.1) | 5 |
| Divalproex sodium | 31 (0.6) | 316 (0.0) | 11 | 14 (0.4) | 21 (0.3) | 2 |
| Gabapentin | 41 (0.8) | 1139 (0.1) | 11 | 24 (0.6) | 49 (0.6) | 0 |
| Lamotrigine | 32 (0.6) | 113 (0.0) | 11 | 18 (0.5) | 18 (0.2) | 5 |
| Phenytoin | 108 (2) | 551 (0.0) | 20 | 48 (1) | 76 (1) | 3 |
| Healthcare contacts, mean±SDj | ||||||
| Number of primary care physician visits | 14.9±19.0 | 9.7±10.6 | 33 | 13.7±17.4 | 13.7±16.2 | 0 |
| Number of internal medicine visits | 3.1±8.5 | 0.8±3.0 | 35 | 2.4±6.9 | 2.3±6.8 | 1 |
| Number of nephrology visits | 0.2±1.3 | 0.1±0.8 | 10 | 0.2±1.1 | 0.2±1.3 | 0 |
| Number of neurology visits | 3.8±5.7 | 0.2±0.9 | 88 | 2.8±4.2 | 2.5±3.8 | 8 |
| Number of psychiatry visits | 0.9±5.6 | 0.6±4.1 | 7 | 1.0±6.1 | 1.0±4.9 | 1 |
| Number of hospitalizations | 1.1±1.6 | 0.4±1.0 | 49 | 0.9±1.4 | 0.9±1.4 | 1 |
| Number of ED visits | 2.2±3.1 | 0.8±1.9 | 54 | 1.9±2.8 | 1.9±3.2 | 2 |
| Healthcare usagek | ||||||
| Serum creatinine tests | 3283 (66) | 1,017,278 (72) | 13 | 2617 (66) | 5275 (66) | 1 |
| CT head | 2362 (47) | 143,204 (10) | 90 | 1620 (41) | 3315 (42) | 2 |
| MRI head | 1781 (36) | 50,669 (4) | 89 | 1164 (29) | 2185 (27) | 5 |
| Electroencephalography | 1702 (34) | 12,678 (0.9) | 97 | 1113 (28) | 2061 (26) | 6 |
| Chest x-ray | 2413 (49) | 445,900 (32) | 35 | 1767 (44) | 3579 (45) | 1 |
| Echocardiography | 1044 (21) | 240,986 (17) | 10 | 770 (19) | 1601 (20) | 2 |
| Epilepsy surgery | 27 (0.5) | 29 (0.0) | 10 | 13 (0.3) | 12 (0.2) | 4 |
| Video EEG monitoring | 365 (7) | 1578 (0.1) | 39 | 209 (5) | 336 (4) | 6 |
| Urinalysis | 1554 (31) | 555,062 (39) | 17 | 1263 (32) | 2493 (31) | 1 |
| Levetiracetam prescriber | ||||||
| Primary care physician | 1749 (35) | — | — | 1435 (36) | — | — |
| Internal medicine | 74 (2) | — | — | 49 (1) | — | — |
| Neurology | 2176 (44) | — | — | 1754 (44) | — | — |
| Neurosurgery | 19 (0.4) | — | — | 11 (0.3) | — | — |
| Psychiatry | 17 (0.3) | — | — | 15 (0.4) | — | — |
| Other/missing | 945 (19) | — | — | 716 (18) | — | — |
| Laboratory data (in subpopulation) | ||||||
| ≥1 serum creatinine measurement | 619 | 202,714 | 509 (13) | 1018 (13) | ||
| Mean serum creatinine level, mg/dl±SD | 0.83±0.29 | 0.93±0.36 | 31 | 0.82±0.27 | 0.84±0.26 | 8 |
| Median serum creatinine level mg/dl (IQR) | 0.77 (0.66–0.94) | 0.86 (0.72–1.04) | 0.77 (0.66–0.93) | 0.79 (0.67–0.96) | ||
| Median eGFR, ml/min per 1.73 m2 (IQR) | 90 (73–105) | 77 (61–89) | 91 (74–108) | 91 (75–105) | ||
| eGFR Category, ml/min per 1.73 m2 | ||||||
| ≥60 | 528 of 619 (85) | 153,348 of 202,714 (76) | 1 | 442 of 509 (87) | 884 of 1018 (87) | 0 |
| 45–60 | 56 of 619 (9) | 29,952 of 202,714 (15) | 8 | 42 of 509 (8) | 84 of 1018 (8) | 0 |
| 15–45 | 35 of 619 (6) | 18,907 of 202,714 (9) | 6 | 25 of 509 (5) | 50 of 1018 (5) | 0 |
| <15 | 0 of 619 (0) | 507 of 202,714 (0.3) | 0 | 0 of 509 (0) | 0 of 1018 (0) | 0 |
| Random urine albumin creatinine ratio, mg/g | ||||||
| <30 | 51 of 75 (68) | 36,665 of 51,154 (72) | 12 | 43 of 58 (74) | 90 of 118 (76) | 0 |
| ≥30 | 24 of 75 (32) | 14,489 of 51,154 (28) | 8 | 15 of 58 (26) | 28 of 118 (24) | 0 |
ACG, Adjusted Clinical Group; TIA, transient ischemic attack; ACE inhibitor, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker; NSAID, nonsteroidal anti-inflammatory drug; ASA, acetylsalicylic acid; ER, emergency department; CT, computed tomography; MRI, magnetic resonance imaging; EEG, electroencephalogram; —, not applicable; IQR, interquartile range.
Data are presented as the number (percentage) of patients, unless otherwise reported.
Standardized differences are less sensitive to sample size than traditional hypothesis tests are. They provide a measure of the difference between groups with respect to the pooled SD; a standardized difference >10% was considered a meaningful difference between the groups.
Income was categorized into fifths of average neighborhood income on the index date.
Comorbid conditions in the 5 yr preceding the index date were considered.
We used the ACG scoring system to score comorbidity. The ACG is a population/patient case-mix adjustment system that provides a relative measure of the individual’s expected consumption of health services. ICD-9/ICD-9-CM codes are categorized into 32 groups, called Ambulatory Diagnostic Groups, on the basis of clinical similarity, chronicity, likelihood of requiring specialty care, and disability. These groups are further reduced to 12 “Collapsed Ambulatory Diagnostic Groups.”
Cancers include lung/bronchi, colon/rectum, breast, pancreas, prostate, leukemia, non-Hodgkin lymphoma, liver, ovarian, and esophageal.
Coronary artery disease includes receipt of coronary artery bypass graft surgery, and percutaneous coronary intervention.
Epilepsy/seizure codes are hospital diagnosis codes and do not capture those patients who do not present to hospital, which underestimates the prevalence of the condition.
Dispensed medications in the 120 d preceding the index date were considered.
Health care contacts in the year preceding the index date were considered.
Health care use in the year preceding the index date was considered.