Association of SSRI and SNRI use with incident hyponatremia after subarachnoid hemorrhage

Abstract

Background:

Hyponatremia is common following subarachnoid hemorrhage (SAH) and is associated with vasospasm and delayed cerebral ischemia (DCI). Risk factors for post-SAH hyponatremia are poorly defined; however, selective serotonin reuptake inhibitors (SSRIs) and selective norepinephrine reuptake inhibitors (SNRIs) are associated with hyponatremia in non-SAH populations. This study assessed whether pre-admission SSRIs/SNRIs were associated with hyponatremia after SAH.

Methods:

This was a single-center retrospective study at a comprehensive stroke center. Patients were included if they were admitted for nontraumatic SAH (aneurysmal, perimesencephalic, or angiography-negative) and were excluded if they died within 24 h of presentation or were hyponatremic on admission. The primary endpoint was the incidence of hyponatremia. Key secondary endpoints included incidence of vasospasm, DCI, and rebleeding and discharge modified Rankin scale.

Results:

292 patients were included; 49 patients were prescribed an SSRI or SNRI prior to admission. Of those included, 55.5 % had a Hunt and Hess score of 1 or 2, and 45 % of patients had anterior aneurysms. 11 % and 26 % of patients had perimesencephalic and CTA-negative SAH respectively. Hyponatremia occurred in 37 % of patients prescribed SSRIs/SNRIs vs 30 % of patients not on SSRIs/SNRIs (p = 0.36). Incidence of radiographic vasospasm in the SSRI/SNRI group was 33 % compared to 21 % in the non-SSRI/SNRI group (p = 0.065).

Conclusions:

We did not find an association that reached our predefined statistical threshold between pre-admission SSRI/SNRI utilization and hyponatremia or secondary endpoints including vasospasm. This study suggests serotonergic therapies do not cause large effects on hyponatremia or clinically relevant complications after SAH.

1. Introduction

Aneurysmal subarachnoid hemorrhage (aSAH) is the leading cause of non-traumatic subarachnoid hemorrhage (SAH). It is characterized by bleeding in the subarachnoid space secondary to a ruptured aneurysm and accounts for approximately 1–4 % of all emergency department visits for acute headache [1–4]. A smaller subset of patients with non-traumatic SAH presents with perimesencephalic SAH (PMSAH) or CTA-negative SAH, which are both associated with improved clinical outcomes and reduced risk of rebleeding compared to aSAH [5]. The incidence of aSAH is higher in women than men, increases with age, and peaks during the fifth decade of life [1]. Patients suffering from aSAH commonly present with severe sudden-onset thunderclap headache, nausea, vomiting, and altered mental status [1,2]. The 30-day mortality of aSAH is estimated to be at least 35 % but is likely underestimated due to undiagnosed patients found dead prior to hospitalization [1,2].

Hyponatremia is a common complication that can develop after SAH. The incidence of clinically significant hyponatremia (defined as serum sodium ≤ 135 mEq/L) reported in the literature is as high as 50 % [6,7]. The mechanism driving hyponatremia is not well understood but generally presents as either syndrome of inappropriate secretion of antidiuretic hormone (SIADH) or cerebral salt wasting syndrome (CSWS) [1]. If left unmanaged or managed inappropriately, hyponatremia is associated with an increased risk of symptomatic DCI [1,6,8–12]. Identification of relevant risk factors for hyponatremia would improve the care of patients with SAH by providing insight into which patients may require increased monitoring.

There are several pharmacologic agents known to contribute to the development of hyponatremia, including both selective serotonin reuptake inhibitors (SSRIs) and selective norepinephrine reuptake inhibitors (SNRIs) [13–15]. Known risk factors for the development of hyponatremia with SSRIs and SNRIs include older age, female gender, concomitant use of diuretics, low body weight, and lower baseline serum creatinine of which several are also connected with risk of SAH [13]. It is not understood whether the utilization of these agents increases the risk of hyponatremia in patients suffering SAH, however. The aim of this study is to evaluate whether pre-admission SSRI/SNRI use is associated with an increased risk of developing clinically significant hyponatremia following SAH.

2. Methods

This single-center, retrospective observational cohort study was conducted at Massachusetts General Hospital (MGH), a comprehensive stroke center, evaluating patients presenting with SAH between January 1, 2017, and January 1, 2022. The study was deemed exempt by the Massachusetts General Institutional Review Board (IRB# 2022P002472). Baseline demographics, admission dates, sodium values, and specific hyponatremia therapy utilization were obtained from the Enterprise Data Warehouse (EDW) and Get with the Guidelines Repository (GWTG). Supplemental manual retrospective chart review was performed to collect pre-admission SSRI/SNRI utilization, Hunt and Hess Scores, WFNS, and the incidence of vasospasm, DCI, and rebleeding based on imaging reports. Patients were considered to have been on an SSRI prior to admission if one of the following medications were on their pre-admission medication list: fluoxetine, citalopram, escitalopram, fluvoxamine, paroxetine, sertraline, vortioxetine. Patients were considered to be on an SNRI if one of the following medications were on their prior to admission medication list: venlafaxine, desvenlafaxine, and duloxetine.

Adult patients ≥ 18 years of age presenting with SAH between January 2017 and January 2022 were screened for inclusion. Patients were included if they presented with SAH observed on computed tomography (CT) imaging or magnetic resonance imaging (MRI) that was then confirmed to be either aneurysmal or non-aneurysmal but non-traumatic (perimesencephalic or angiography-negative) by CT angiography (CTA) or digital subtraction angiography (DSA). Patients were excluded if they died within 24 h of admission, were hyponatremic on admission (serum sodium < 135 mEq/L) or had SAH secondary to reversible cerebral vasoconstriction syndrome (RCVS) or cerebral amyloid angiopathy (CAA).

The primary endpoint of this study was the incidence of hyponatremia, which was defined as a serum sodium < 135 mEq/L on two consecutive basic metabolic panels between patients with and without pre-admission SSRI/SNRI utilization. Secondary endpoints included incidence of rebleeding or aneurysm expansion on imaging, radiographic vasospasm, and DCI. Incidence of vasospasm was captured from radiographic imaging including CTA, magnetic resonance angiography (MRA), or DSA reports. Changes in neurologic exam or transcranial doppler (TCD) values were not included due to their subjectivity. Incidence of DCI was defined as a new infarct seen on CT or MRI 48 h or more after the index bleed [16]. Differences in discharge functional status as assessed on the modified Rankin scale (mRS) were also evaluated between groups, as well as time to hyponatremia and time to vasospasm. Additional endpoints included identification of other potential risk factors for hyponatremia and rates of hyponatremia therapy utilization (fludrocortisone, normal saline infusions, hypertonic saline).

Continuous and ordinal variables were compared between groups using the Mann-Whitney U-test or Student’s t-test as appropriate. Categorical variables between groups were compared utilizing Chi-squared or Fisher’s exact test as appropriate. Continuous variables were summarized using medians and interquartile range and significance was defined as an alpha level of 0.05. The adjusted association with pre-admission SSRI/SNRI use with hyponatremia was assessed using multivariable logistic regression, adjusting for age, Hunt and Hess and World Federation of Neurosurgical Societies scores, aneurysm location, past medical history, and smoking history. Time to incident hyponatremia and vasospasm was assessed with Kaplan Meier survival analysis. Data were analyzed utilizing R, version 4.2.1 (R Core Team, R Foundation for Statistical Computing, Vienna, Austria.)

3. Results

A total of 292 patients were included in the final analysis. Among them, 49 (16.7 %) patients were prescribed an SSRI or SNRI prior to admission, and 243 (83.2 %) were not. The most common reasons for exclusion included presentation for RCVS, hyponatremia on admission, and mortality within 24 h of presentation (Fig. 1). The two groups were well balanced regarding baseline characteristics (Table 1).

Fig. 1.

CONSORT diagram for patient inclusion and exclusion.

Table 1.

Baseline characteristics of included population.

Characteristic	Not on SSRI or SNRI, N = 243*	On SSRI or SNRI, N = 49*
Age (years)	59 (49, 67)	61 (51, 68)
Male	104 (43 %)	14 (29 %)
Race
White	168 (69 %)	37 (76 %)
Black or African	15 (6.2 %)	3 (6.1 %)
American
Asian	11 (4.5 %)	4 (8.2 %)
Unavailable	49 (20.2 %)	5 (10.2 %)
BMI	27 (24, 32)	28 (24, 32)
Diabetes	22 (9.1 %)	9 (18 %)
Heart Failure	3 (1.2 %)	2 (4.1 %)
Smoking History	53 (22 %)	8 (16 %)
Hypertension	97 (40 %)	26 (53 %)
Hunt and Hess Scale
I	54 (22 %)	9 (18 %)
II	83 (34 %)	16 (33 %)
III	48 (20 %)	9 (18 %)
IV	23 (9.5 %)	8 (16 %)
V	35 (14 %)	7 (14 %)
WFNS Scale
1	115 (47 %)	23 (47 %)
2	35 (14 %)	7 (14 %)
3	15 (6.2 %)	2 (4.1 %)
4	40 (16 %)	6 (12 %)
5	38 (16 %)	11 (22 %)
Aneurysm Location
Anterior^	105 (43 %)	27 (55 %)
Posterior†	40 (16 %)	10 (20 %)
CTA Negative	65 (27 %)	11 (22 %)
Perimesencephalic	33 (14 %)	1 (2.0 %)
Aneurysm Size (mm)	6.0 (4.0, 8.0)	5.0 (3.0, 7.0)
EVD Placement	125 (51 %)	23 (47 %)

^*n (%);

^{^}Aneurysms in internal carotid artery (ICA), middle cerebral artery (MCA), anterior cerebral artery (ACA), or anterior communicating artery (AComm) made up 72.4 % of all identified aneurysms;

^†Aneurysms in the vertebral artery, basilar artery, posterior cerebral artery (PCA), or any cerebellar artery made up 27.5 % of all identified aneurysms

The median age was 60, 59.6 % of patients were female, and 70 % of patients were white. Comorbid conditions were similar between groups, with hypertension being the most prevalent in 42 % of patients. SAH severity based on Hunt and Hess Scores as well as World Federation of Neurological Surgeons score (WFNS) was comparable between groups. Most patients had aSAH (62.3 %) of which 72.5 % of aneurysms were in the anterior circulation and 27.4 % were located in the posterior circulation. CTA negative SAH accounted for 26 % of patients and 11.6 % of patients had SAH in a perimesencephalic pattern. The aneurysm size was also similar between the cohorts (5.0 mm in the SSRI/SNRI group vs. 6.0 mm in the non-SSRI/SNRI group) Around half of patients had an extra-ventricular drain (EVD) placed in each group.

The primary outcome of hyponatremia occurred in 37 % of patients prescribed SSRIs or SNRIs compared to 30 % in the patients not prescribed these medications prior to admission (OR 1.35; 95 % CI 0.70–2.55). The time to incident hyponatremia was similar between groups (6.0 days vs. 5.5 days; p = 0.80) and appeared to plateau after 14 days (Fig. 2). There were no significant differences in rates of rebleeding (5.4 % vs. 4.1 %; p > 0.99) or DCI (14 % vs. 12 %, p = 0.80) between the groups (Table 2). Radiographic vasospasm occurred more frequently in the SSRI/SNRI cohort, but did not reach the predefined threshold for statistical significance (33 % vs 21 %; p = 0.065). The incidence of specific hyponatremia therapies was consistent between the two cohorts (Table 3), with most patients receiving treatment with hypertonic saline 3 % infusions.

Fig. 2.

Time to hyponatremia in the SSRI/SNRI cohort and non-SSRI/SNRI cohort.

Table 2.

Primary and secondary clinical outcomes.

Characteristic	Not on SSRI or SNRI, N = 243*	On SSRI or SNRI, N = 49*	p-value†
Hyponatremia	73 (30 %)	18 (37 %)	0.36
Rebleeding	13 (5.4 %)	2 (4.1 %)	> 0.99
Delayed Cerebral Ischemia	33 (14 %)	6 (12 %)	0.80
Vasospasm	50 (21 %)	16 (33 %)	0.065

^*n (%)

^†Pearson’s Chi-squared test; Fisher’s exact test

Table 3.

Pharmacologic therapy utilization for treatment of hyponatremia.

Characteristic	Not on SSRI or SNRI, N = 243*	On SSRI or SNRI, N = 49*	p-value†
Hyponatremic	73 (30 %)	18 (37 %)	0.36
Fludrocortisone	34 (14 %)	7 (14 %)	0.96
Hypertonic Saline Infusion	68 (28 %)	15 (31 %)	0.71
Hypertonic Saline Bolus	24 (9.9 %)	4 (8.2 %)	> 0.99
Sodium chloride tablets	36 (15 %)	10 (20 %)	0.33

^*n (%)

^†Pearson’s Chi-squared test; Fisher’s exact test

There was not a global difference in the distribution of discharge mRS scores between groups (OR 1.29, 95 % CI 0.76–2.16). However, patients taking SSRIs/SNRIs prior to admission were less likely to be discharged with an mRS of 0–2 (OR 0.46; 95 % CI 0.22–0.91).

Secondary analyses completed included an exploratory analysis evaluating the incidence of hyponatremia based on the definition of one sodium value on BMP < 135 mEq which did not demonstrate a significant difference in hyponatremia (53 % vs. 51 %; p = 0.75). A multivariable analysis performed to assess the cohort for alternative risk factors associated with the development of hyponatremia revealed hyponatremia was more commonly observed in patients with a Hunt and Hess scale of 3 or 4 and prolonged length of stay (Table 4). An additional subgroup analysis was performed comparing outcomes based on aneurysm location that revealed a significant difference in discharge mRS in patients presenting with anterior SAH that were on SSRIs/SNRIs prior to admission and those that were not (Table 5). A discharge mRS score of 0 was only found in patients that presented with CTA-negative and perimesencephalic SAH.

Table 4.

Predictors of the primary endpoint (hyponatremia).

Characteristic	OR*	95 % CI*
Age (per year)	1.00	0.91, 1.01
Gender, Male	0.86	0.52, 1.43
Past Medical History
Diabetes	1.46	0.66, 3.11
Heart Failure	1.48	0.19, 9.10
Smoking History	1.00	0.53, 1.82
Hypertension	1.27	0.77, 2.09
Hunt and Hess Scale
1	—	——
2	2.87	1.30, 6.85
3	3.50	1.48, 8.86
4	6.40	2.42, 18.0
5	2.69	1.04, 7.25
WFNS Scale
1	——	——
2	2.45	1.16, 5.11
3	4.05	1.43, 11.7
4	1.74	0.82, 3.62
5	2.48	1.23, 5.00
Aneurysm Location
Anterior	—	—
Posterior	1.59	0.82, 3.09
CTA Negative	0.54	0.28, 1.02
Perimesencephalic	0.32	0.10, 0.82
Aneurysm Size (per mm)	1.00	0.93, 1.07
Hospital Length of Stay	1.05	1.03, 1.07

^*OR = Odds Ratio, CI = Confidence Interval

Table 5.

Subgroup Analyses of discharge mRS, incidence of rebleeding, vasospasm, and delayed cerebral ischemia based on aneurysm location.

	Anterior			Posterior			CTA Negative			Perimesencephalic
	Not on SSRI/SNRI N = 105a	On SSRI/SNRI N = 27a	p-valueb	Not on SSRI/SNRI N = 40a	On SSRI/SNRI N = 10a	p-valuec	Not on SSRI/SNRI N = 65a	On SSRI/SNRI N = 11a	p-valuec	Not on SSRI/SNRI N = 33a	On SSRI/SNRI N = 1a	p-valuec
Discharge			0.03			0.4			0.3			0.3
mRS
1	22 (21 %)	2 (7.7 %)		4 (10 %)	0 (0 %)		16 (25 %)	5 (45 %)		13 (39 %)	0 (0 %)
2	18 (17 %)	4 (15 %)		3 (7.5 %)	0 (0 %)		4 (6.2 %)	1 (9.1 %)		12 (36 %)	0 (0 %)
3	11 (11 %)	9 (35 %)		6 (15 %)	3 (30 %)		11 (17 %)	3 (27 %)
4	21 (20 %)	3 (12 %)		15 (38 %)	3 (30 %)		15 (23 %)	0 (0 %)		3 (9.1 %)	1 (100 %)
5	10 (9.7 %)	5 (19 %)		7 (18 %)	4 (40 %)		7 (11 %)	2 (18 %)
6	21 (20 %)	3 (12 %)		5 (12 %)	0 (0 %)		7 (11 %)	0 (0 %)		1 (3.0 %)	0 (0 %)
Hyponatremic	46 (35 %)	27 (44 %)	0.8	23 (46 %)	32 (61 %)	0.7	17 (22 %)	14 (34 %)	0.7	5 (15 %)	4 (12 %)
Rebleeding	9 (8.7 %)	2 (7.4 %)	> 0.9	3 (7.7 %)	0 (0 %)	> 0.9	1 (1.5 %)	0 (0 %)	> 0.9	0 (0 %)	0 (0 %)
Vasospasm	28 (27 %)	10 (37 %)	0.3	14 (35 %)	5 (50 %)	0.5	8 (12 %)	1 (9.1 %)	> 0.9	0 (0 %)	0 (0 %)
DCI	20 (19 %)	4 (15 %)	0.8	5 (12 %)	2 (20 %)	0.6	8 (12 %)	0 (0 %)	0.6	0 (0 %)	0 (0 %)

^an (%)

^bFisher’s exact test; Pearson’s Chi-squared test

^cFisher’s exact test

4. Discussion

Hyponatremia is the most common electrolyte abnormality observed following SAH and may lead to various neurologic complications including altered mental status, seizures, and coma [6,8]. The etiology of hyponatremia in this patient population is diverse; however, it is most commonly attributed to SIADH. SIADH is associated with excessive secretion of antidiuretic hormone (ADH) secondary to simulation of the hypothalamus from traumatic or ischemic factors, resulting in enhanced water reabsorption, fluid retention, and ultimately dilutional hyponatremia [17]. SIADH is also a known adverse effect of several medication classes including SSRIs and SNRIs. The mechanism by which SSRIs/SNRIs lead to SIADH is through inhibition of norepinephrine reuptake which causes simulation of ADH release [18]. Therefore, we proposed a theoretical increased risk of hyponatremia in patients with SAH secondary to both hypothalamic stimulation and increased norepinephrine causing increased secretion of ADH.

Identifying risk factors for the development of hyponatremia following SAH would provide framework for selective monitoring based on overall risk stratification. This single-center retrospective cohort study of patients with SAH did not reveal an increased risk of hyponatremia in patients on pre-admission SSRIs or SNRIs. Despite a known correlation between SSRIs/SNRIs and SIADH, their utilization did not appear to be associated with an increased incidence of hyponatremia in hospitalized patients with SAH [13]. Similar results were revealed for rates of rebleeding, vasospasm, and DCI, with similar rates of clinically relevant outcomes between groups. In addition, the distribution of discharge mRS was similar between the two groups, further supporting the safety of SSRIs and SNRIs prior to SAH (Fig. 3).

Fig. 3.

Distribution of Modified Rankin Score (mRS) at Discharge in the SSRI/SNRI Cohort and Non-SSRI/SNRI Cohort (OR 1.29, 95 % CI 0.76–2.16).

To our knowledge, this is the first evaluation of the risk of hyponatremia in patients taking SSRIs or SNRIs that experienced non-traumatic SAH. There is mixed evidence on the effect of SSRI/SNRIs on clinically relevant outcomes in patients with SAH. A 2005 retrospective study by Singhal et. al included 514 patients with SAH and assessed the incidence of vasospasm with several pharmacologic agents [19]. Their study found that pre-admission SSRI utilization was associated with both asymptomatic and symptomatic vasospasm (19.2 % vs. 5.1 %; p = 0.019) [19]. These results are consistent with the trend found in this cohort, as patients on SSRIs/SNRIs demonstrated higher rates of radiographic vasospasm, although we were underpowered to reach the threshold for statistical significance. Alternatively, a 2016 study by Young et. al evaluated a larger cohort of 579 patients and found that SSRIs use prior to and during hospitalization for aSAH was not associated with radiological infarction, DCI, symptomatic vasospasm, or functional outcomes at 3 months [20]. Both studies by Young and Singhal defined vasospasm as either symptomatic or radiographic and included patients with focal neurologic defects regardless of radiographic evidence of vasospasm as well as patients with vasospasm as diagnosed by TCD changes [19,20]. The methodology differs from our study, which only included patients with documented vasospasm on CTA or MRA. Additionally, neither aforementioned studies included patients with perimesencephalic or CT-negative SAH, both of which were included in this analysis [19,20]. Neither study included patients presenting with RCVS, which is associated with increased incidence of reversible vasospasm, especially with SSRI/SNRI use. Only 7.1 % of the patients included in Singhal et. al’s study population were prescribed SSRIs/SNRIs prior to admission, which weakens the overall quality of evidence when compared to the larger cohort of patients seen in the study completed by Young [19,20]. With the ambiguity of results based on small cohorts, the findings of this study must be interpreted cautiously, and do not yet indicate clinical interventions. There remains a theoretical serotonergic mechanism associated with vasospasm risk; however, larger, powered studies are required to further characterize this relationship and guide any potential clinical practice change (Fig. 4).

Fig. 4.

Time to Vasospasm in the SSRI/SNRI Cohort and Non-SSRI/SNRI Cohort.

Amongst the secondary outcomes evaluated, hyponatremia appeared to be more commonly observed in patients with higher grade WFNS as well as Hunt & Hess score in addition to patients with prolonged length of stay. These findings are consistent with previous literature, as length of stay has consistently been found to be prolonged in patients that experience hyponatremia, though causality of the prolonged length of stay is more likely secondary to the development and management of hyponatremia [6,10–12,15,21]. An increased risk of hyponatremia was not observed with baseline characteristics including a diagnosis of heart failure. There were no differences in rates of hyponatremia relating to the location of aneurysm; however, aneurysm location was associated with differences in discharge mRS. The only patients with discharge mRS of 0 had either CTA-negative or perimesencephalic SAH. Interestingly, in patients with anterior subarachnoid hemorrhage, patients on SSRIs/SNRIs were more likely to have higher discharge mRS. This finding may warrant additional investigation into the impacts of serotonergic agents on discharge mRS in specific aneurysm locations.

A notable finding of this study was the apparent plateau of hyponatremia incidence in both groups around 14 days (Fig. 2). To compare, another observational study by Ridwan et. al in Germany evaluated the incidence of hyponatremia following aSAH at day 1, 3, 7–10, and 14–21 following presentation and found that the highest incidence of hyponatremia occurred at day one and at day 7–10 [7]. The aforementioned study included patients that were hyponatremic on admission, which creates a dichotomy from our findings, as we excluded patients with hyponatremia upon presentation. Despite this, the results of this study in addition to ours suggest a “hyponatremia window” akin to the vasospasm window with the highest risk period being the first 14 days post-ictus.

There are several notable limitations to this study. First, confirmation of pre-admission SSRI/SNRI use relied on accurate medication reconciliation done at the time of admission and was collected retrospectively based on chart documentation. Medication histories are documented on patients on admission; however, error is possible which may contribute to information bias in our findings. Additionally, medication lists utilized do not account for potential poor adherence or differences in duration of therapy prior to presentation. Additionally, SSRI/SNRI dosages and duration of therapy were not collected, which would have provided valuable insight on the differentiation of effect associated with higher doses and longer treatment duration on the evaluated endpoints. This study evaluated the incidence of hyponatremia relevant to pre-admission SSRI/SNRI utilization; however, the continuation of these medications while admitted was not collected. Stratification of incident hyponatremia based on continuation of SSRIs/SNRIs once admitted may have contributed to a larger global evaluation and should be considered in future studies.

Secondly, there are potentially confounding variables that were not accounted for that may also have contributed to the rates of hyponatremia seen in this cohort which could affect the relationship between SSRI/SNRIs and hyponatremia. For example, alternative medications implicated in causing hyponatremia such as thiazide-type diuretics, vasopressin, and anti-epileptics such as oxcarbazepine were not collected. Another limitation is the lack of differentiation completed between the etiologies of hyponatremia experienced. An important consideration of our findings relates to the practice culture at our institution, in which, at early evidence of hyponatremia, patients are often aggressively treated and corrected with pharmacologic intervention. This practice pattern may have falsely minimized the actual number of patients that experienced hyponatremia in our cohort. To account for this, a secondary analysis was completed including patients with one sodium value < 135 mEq/L. The results of this secondary analysis are consistent with our primary analysis demonstrating an insignificant difference between the groups in both rates of hyponatremia as well as secondary outcomes. Additionally, vasospasm was defined radiographically, and asymptomatic vasospasm may have been included in our cohort.

Another limitation of our findings includes the methodology in which treatment modalities including surgical procedures such as craniotomy or endovascular intervention were not evaluated. These interventions may also have had an impact on the findings that our study has not accounted for. Another limitation is that Modified Fisher Grading Scale values were not collected. This information would have provided additional context on the illness severity of the patients included in the study. A final limitation is the retrospective nature of the evaluation, as the quality of data is reliant on the data entered at the time of patient presentation.

5. Conclusions

This study does not support increased risk of hyponatremia in patients on pre-admission SSRIs and SNRIs. SSRIs and SNRIs were also not associated with increased risk of clinically relevant complications of SAH, including rebleeding or DCI, although a trend towards increased incidence of radiographic vasospasm was observed. Additional larger studies are required to further characterize the relationship between SSRI/SNRI utilization and patient outcomes in SAH.

Abbreviations:

ASAH

aneurysmal subarachnoid hemorrhage

CAA

cerebral amyloid angiopathy

CSWS

cerebral salt wasting syndrome

CTA

computed tomography angiography

DCI

delayed cerebral ischemia

DSA

digital subtraction angiography

EDW

enterprise data warehouse

RCVS

reversible cerebral vasoconstriction syndrome

SIADH

syndrome of inappropriate antidiuretic hormone

SNRI

selective norepinephrine reuptake inhibitor

SSRI

selective serotonin reuptake inhibitor

TCD

transcranial doppler

WFNS

World Federation of Neurological Surgeons