Patient and treatment demographics
A total of 623 patients (median age: 75 [68–81] years, male/female ratio: 2.24) who underwent surgical cSDH evacuation during the study period were included (Fig.
1). Headaches (28%), coordination deficits (24%), aphasia (11%), and confusion (11%) were the most common initial symptoms that patients reported. The most common leading symptoms at admission were coordination deficits (26%), hemiparesis (15%), aphasia (14%), and reduced vigilance (14%). The prevalence of comorbidities in the cSDH patient population was high (63%), and arterial hypertension, as well as cardiac arrhythmias, accounted for most of them (35% and 21%, respectively). On head CT imaging, cSDH had a median thickness of 21 [15-25] mm, were right-sided in 31%, left-sided in 41%, and bilateral in 28% of cases, while a midline shift was visible in 67% with a median of 6 [2-10] mm. Antithrombotics were present in more than half of all cSDH patients (51%) and were almost equally distributed between only anticoagulant (47%) and only antiplatelet (50%) drugs, with few patients (3%) receiving both. Substances used for anticoagulation were either VKA (89%), new oral anticoagulants (10%), or heparin (1%). Inhibition of platelet function was mostly induced with acetylsalicylic acid (88%) or with clopidogrel (3%) and sometimes with a combination of both drugs (9%).
Management of antithrombotics included assessment of plasmatic coagulation or platelet function, which influenced the timing of surgery: Provided that no urgent indication was present, surgery was delayed for a median of 3 [2-4] days post-admission with the antithrombotics being terminated and the corresponding patients (29%) being clinically observed on the normal ward. Moreover, hemostatic therapy was applied in 55% of all such patients before or during surgery. Thereby, PCC (62%), phytomenadione (7%; Konakion®, Roche Pharmaceuticals, Grenzach, Germany) or PCC, and phytomenadione (19%) and in rare cases, fresh frozen plasma (FFP; 3%), were used for anticoagulants and desmopressin (55%; Minirin®, Ferring Arzneimittel, Kiel, Germany), tranexamic acid (18%; Cyklokapron®, Pfizer, Berlin, Germany), a combination of both drugs (10%), or platelet transfusions (12%) for antiplatelets.
Hematoma evacuation was performed by burr hole drainage in 93% of cases, while the remaining patients received a craniotomy. One or more subdural drains were placed in all patients. During the hospital stay, nine patients (1%) died from inter alia cardiopulmonary insufficiency, multiorgan dysfunction due to sepsis, or acute secondary hemorrhage with trans tentorial herniation. On discharge, 94% of the remaining patients had a GCS score between 13 and 15, and the median GOS and mRS scores were 5 [4-5] and 1 [1-3], respectively. While half of the patients needed secondary hospital or rehabilitation care, the other half could be discharged home. Within 30 days of the primary surgical intervention, the overall reoperation rate due to remaining or recurrent cSDH was 23%.
Reoperation after cSDH recurrence
Reoperation within 30 days of the primary hematoma evacuation was performed in 145 cSDH patients (23%) due to either neurological deterioration or missing improvements caused by residual or recurrent hematomas. A burr hole trepanation was performed in 58% of those cases, while 42% of patients underwent a craniotomy during the second operation.
In the univariate logistic analysis, only the presence of comorbidities in general, as well as arterial hypertension and renal insufficiency as comorbidities, were found to be significant predictors for the need for reoperation within 30 days (Table
1). Moreover, in the multivariate analysis that included age, gender, known comorbidities, antithrombotic medication, and cSDH thickness as covariates, only a history of chronic comorbidity (OR 2.12; 95% CI, 1.30–3.55;
p = 0.003) was found to be an independent predictor for reoperation as well.
Table 1
Univariate and multivariate analyses of predictive factors for reoperation due to cSDH recurrence
No. of patients | 478 | 145 | - | - | - |
Gender | | | 0.581 | 0.504 | 1.18 (0.72–1.90) |
Male | 328 (69%) | 103 (71%) | | | |
Female | 150 (31%) | 42 (29%) | | | |
Median age [IQR] | 75 [68–82] years | 75 [69–80] years | 0.482 | 0.192 | 0.99 (0.97–1.00) |
Antithrombotics | 234 (44%) | 84 (54%) | 0.059 | 0.163 | 1.20 (0.81–1.77) |
Anticoagulants | 109 (23%) | 42 (29%) | 0.193 | | |
Antiplatelets | 117 (24%) | 42 (29%) | 0.384 | | |
Both | 8 (2%) | 0 (0%) | 0.076 | | |
Comorbidities |
Known comorbidities | 280 (59%) | 111 (77%) | p < 0.001 | 0.003 | 2.12 (1.30–3.55) |
Arterial hypertension | 152 (32%) | 63 (43%) | 0.010 |
Cardiac arrhythmias | 90 (19%) | 38 (26%) | 0.055 |
Coronary artery disease | 53 (11%) | 25 (17%) | 0.052 |
Stroke history | 26 (5%) | 9 (6%) | 1.000 |
Diabetes mellitus | 63 (13%) | 24 (17%) | 0.306 |
Malignancy | 41 (9%) | 14 (10%) | 0.689 |
Renal insufficiency | 27 (6%) | 16 (11%) | 0.028 |
Alcohol abuse | 11 (2%) | 4 (3%) | 0.753 |
Median laboratory values [IQR] |
INR | 1.03 [0.87–1.09] | 1.04 [0.88–1.14] | 0.768 | | |
aPTT (s) | 25.1 [23.3–27.3] | 24.9 [23.4–27.5] | 0.981 | | |
Platelet count (109/L) | 243 [194–303] | 242 [193–321] | 0.458 | | |
Creatinine (mg/dL) | 0.84 [0.73–1.02] | 0.85 [0.70–1.07] | 0.219 | | |
GFR (mL/min/1.73m2) | 81 [66–92] | 79 [63–92] | 0.568 | | |
Initial clinical presentation |
GCS [IQR] | 15 [14–15] | 14 [14–15] | 0.268 | | |
mRS [IQR] | 2 [2–3] | 3 [2–3] | 0.095 | | |
Midline shift | 330 (82%) | 89 (79%) | 0.595 | | |
Median cSDH thickness [IQR] | 20 [15–25] mm | 22 [16–27] mm | 0.238 | 0.177 | 1.02 (0.99–1.05) |
Patients with anticoagulants, antiplatelets, and without antithrombotics
In our study, 151 patients were on anticoagulant and 159 patients were on antiplatelet medication, respectively (Table
2). Compared to patients without antithrombotic medication, those patients were significantly older (
p < 0.001 each) and predominately males (
p = 0.0432 each; Supplement
1).
Table 2
Characteristics, clinical course, and outcome in cSDH patients with anticoagulant, antiplatelet, and no antithrombotic medication
No. of patients | 305 | 151 | 159 | - |
Gender | 0.006 |
Male | 193 (63%) | 114 (76%) | 119 (75%) | |
Female | 112 (37%) | 37 (24%) | 40 (25%) | |
Median age [IQR] | 73 [63–79] years | 77 [71–80] years | 77 [70–84| years | < 0.001 |
Initial clinical presentation |
GCS 13–15 | 272 (89%) | 129 (85%) | 140 (88%) | 0.428 |
GCS 9–12 | 17 (6%) | 16 (11%) | 13 (8%) | 0.150 |
GCS < 9 | 13 (4%) | 5 (3%) | 5 (3%) | 0.788 |
mRS [IQR] | 2 [2, 3] | 3 [2, 3] | 3 [2, 3] | 0.164 |
Comorbidities |
Known comorbidities | 155 (51%) | 125 (83%) | 103 (65%) | < 0.001 |
Arterial hypertension | 79 (26%) | 66 (44%) | 66 (42%) | < 0.001 |
Cardiac arrhythmias | 21 (7%) | 90 (60%) | 11 (7%) | < 0.001 |
Coronary heart disease | 18 (6%) | 26 (17%) | 31 (19%) | < 0.001 |
Stroke history | 6 (2%) | 13 (9%) | 15 (9%) | < 0.001 |
Diabetes mellitus | 25 (8%) | 26 (17%) | 35 (16%) | < 0.001 |
Renal insufficiency | 10 (3%) | 17 (11%) | 15 (9%) | 0.002 |
Alcohol abuse | 11 (4%) | 3 (2%) | 1 (1%) | 0.131 |
Complications |
Complications: cardiovascular | 4 (1%) | 10 (7%) | 3 (2%) | 0.003 |
Complications: pulmonary | 11 (4%) | 8 (5%) | 12 (8%) | 0.181 |
Complications: coagulation | 5 (2%) | 8 (5%) | 0 (0%) | 0.004 |
Complications: neurological | 46 (15%) | 35 (23%) | 39 (25%) | 0.022 |
Outcome |
In-hospital mortality | 5 (2%) | 1 (1%) | 3 (2%) | 0.628 |
GOS at discharge [IQR] | 5 [5–5] | 5 [4–5] | 5 [4–5] | 0.595 |
mRS at discharge [IQR] | 1 [1–3] | 2 [1–3] | 2 [1–3] | 0.262 |
Reoperation < 30 days | 61 (20%) | 42 (28%) | 42 (26%) | 0.112 |
A mild brain injury (GCS 13–15) was documented in 85% and 88% of patients on anticoagulants and antiplatelets, respectively, vs. in 89% of patients without antithrombotics. A moderate injury (GCS 9–12) was present in 11% (anticoagulant) and 8% (antiplatelet) vs. 6% (no antithrombotic medication) and a severe injury (GCS < 9) in 3% (anticoagulant) and 3% (antiplatelet) vs. 4% (no antithrombotic medication).
Known comorbidities were most frequently found in patients on anticoagulants (83%), and thus significantly more common than in patients on antiplatelets (65%;
p = 0.001; Supplement
1). Compared to patients without antithrombotic medication, however, known comorbidities (51%) were found significantly more often in both, the anticoagulants (
p < 0.001) and the antiplatelets (
p = 0.010) groups. As expected, cardiac arrhythmias were predominantly found in patients on anticoagulants (60%), while patients on antiplatelets most often suffered from arterial hypertension (42%). Furthermore, anticoagulant and antiplatelet medication were both significantly associated with the comorbidities stroke (
p = 0.005 and
p = 0.003), diabetes mellitus (
p = 0.018 and
p < 0.001), and renal insufficiency (
p = 0.008 and
p = 0.028). Significant inter-group differences could also be observed in the occurrence of cardiovascular and coagulative complications, which were most frequently found in patients on anticoagulants (7% and 5%, respectively), while neurological complications were highest in the antiplatelets group (25%; Table
2). A significantly higher burden of cardiovascular complications was documented in patients on anticoagulants compared to patients without antithrombotic medication (
p = 0.028; Supplement
1).
There were, however, no significant differences regarding in-hospital mortality, GOS, mRS, or reoperation risk within 30 days between patients on anticoagulants, on antiplatelets, or without antithrombotic medication.