Absence of Longer Reorientation Times in Patients Undergoing Electroconvulsive Therapy and Concomitant Treatment with Lithium.

OBJECTIVE: Lithium is a drug of choice in the treatment of bipolar disorder and refractory depressive disorders. However, previous research suggests lithium has a negative cognitive impact in recovery from electroconvulsive therapy (ECT) and a higher risk of delirium, so patients are often required to stop taking lithium before ECT, despite risk of relapse. We studied the cognitive impact of serum lithium levels in patients undergoing ECT.
METHODS: This was an observational prospective study. Serum lithium levels, thyroid and biochemical parameters were measured prior to each ECT session. Time elapsed from the anesthetic induction to the electrical stimulus and then to the patients' reorientation was recorded, as well as the motor seizure duration and electroencephalogram (EEG) seizure duration. A statistical analysis using a linear mixed model was run while adjusting for confounding factors.
RESULTS: Ten participants underwent a total of 86 ECT sessions (41% right unilateral ultrabrief pulse, and 59% bilateral brief pulse). A negative interaction between lithium levels and reorientation time was found among those doing bilateral brief pulse ECT. No association was observed in patients doing unilateral ultrabrief pulse ECT. No significant relationship was observed between lithium and both motor and EEG-assessed seizure duration.
CONCLUSION: This study suggests that low to moderate serum lithium levels (< 0.7 mmol/L) might have no harmful cognitive effects in patients under right unilateral ultrabrief pulse and bilateral brief pulse ECT.

INTRODUCTION

Electroconvulsive therapy (ECT) is recommended for acute and maintenance treatment of mood and psychotic refractory disorders, as well as in acute emergent settings. There is also strong evidence supporting its robust antisuicidal effect [1]. Its main adverse effect consists of cognitive impairment, such as acute transient disorientation, anterograde and retrograde amnesia [1]. As for lithium carbonate, it is the most widely studied mood stabilizer and a well-known first line acute and maintenance treatment for bipolar disorders [2,3] as well as an adjunctive pharmacological option for resistant depression [4]. It is also the mainstay of treatment for preventing relapse after an acute cycle of ECT. Its toxicity comprises a wide variety of symptoms, ranging from nausea and vomiting when mild, to coma, convulsions and collapse when severe [5]. Concomitant therapy with lithium and ECT is not infrequent: patients with unipolar depression or bipolar disorder medicated with lithium might be referred to ECT and, in patients for whom ECT was effective, lithium is strongly regarded as maintenance treatment option to prevent relapse after ECT [2].

The literature is vast respecting both lithium and ECT individually, though still not as detailed when concerning their combined effects in terms of safety and risks regarding patients’ neurocognitive function. Evidence in this matter consists mostly of case reports [6-17] as well as a few retrospective and naturalistic prospective studies. There is only one prospective controlled study regarding concomitant use of lithium and ECT [18]. Recently a study has concluded that concomitant therapy is associated with a higher incidence of both acute delirium and transient cognitive impairment when compared to ECT alone [19]. Due to the reduced amount of studies addressing this matter, both the Royal College of Psychiatrists [20] and the American Psychiatric Association’s [21] guidelines advise for the withdrawal of lithium, or at least lowering its concentration to subtherapeutic levels, prior to starting ECT. Although some more evidence is starting to emerge regarding the interaction between lithium and ECT, the path is still unclear as sometimes important confounders are not taken into account. We still lack certainty about which aspects influence post-ECT recovery, especially when the patient is also medicated with lithium carbonate. This leaves clinical practice with difficult decision-making situations. Should lithium be stopped at all before ECT (e.g., in a suicidal patient)? Should the dosage be reduced? So, the recommendation to simply stop lithium seems sometimes detached from clinical practice and the need to deal with difficult-to-treat patients. The future of medicine and psychiatry is moving towards personalized care, which might mean, in the case of lithium and ECT, making fine decisions depending on clinical severity, previous lithium response, lithium and ECT effectiveness, side effects during each session, monitoring lithium levels, and also considering patient preferences.

This study aimed to assess the relationship between serum lithium levels and the reorientation time of patients under combination therapy while adjusting for potential confounders such as age, sex, type of ECT (right unilateral ultrabrief pulse or bilateral brief pulse), anesthetic drug used, time interval between anesthetic induction and electrical discharge, charge, electroencephalogram (EEG) seizure duration, motor seizure duration, titration session, thyroid stimulating hormone (TSH), triiodothyronine (T3), thyroxine (T4), sodium (Na＋), potassium (K＋), and chlorine (Cl–) levels.

METHODS

Setting and Subjects

The study took place at the ECT Unit of the Department of Psychiatry and Mental Health, Cova da Beira University Hospital Center (Translated from the its official name: Centro Hospitalar e Universitário da Cova da Beira, Covilhã, Portugal). ECT was offered both to inpatients and outpatients from the hospital’s catchment area as well as from other hospitals not able to provide ECT.

All patients medicated with lithium therapy while undergoing ECT were recruited between May 2019 and January 2020 according to the study’s selection criteria. These criteria were reviewed based on the patient’s clinical records and charts. Selection criteria were: currently or about to enroll in ECT, age between 18 and 80 years, established International Classification of Diseases-10 or 11 diagnosis of major depressive disorder, bipolar disorder or other affective disorder, no use of non-steroidal anti-inflammatory drugs (NSAIDs), angiotensin-converting enzyme (ACE) inhibitors or calcium channel blockers within the last 48 hours (given their risk of increasing serum lithium levels) [5], no excessive alcohol consumption nor recreational drug addiction, a mini-Mental Exam score of 22 or greater prior to ECT, absence of any sort of untreated thyroid or parathyroid dysfunction, uncontrolled hypertension, osteoporosis, previous diagnosis of diabetes insipidus, not undergoing treatment with Cytochrome P450 inducing (such as carbamazepine or rifampicin), or inhibiting drugs (such as amiodarone, spironolactone, or ethinyl estradiol). Were excluded from the study patients who had a history of previously experiencing life threatening effects from either ECT or lithium carbonate intake (acute lithium toxicity, intracranial hypotension due to low cerebrospinal fluid [CSF] volume, or anesthesia-related adverse effects), renal impairment with glomerular filtration rate of ≤ 45 ml/min/1.73 m², and chronic users of NSAIDs, ACE inhibitors, or calcium channel blockers.

ECT Procedure

At the ECT Unit where the study was performed, right unilateral electrode placement with 0.3 ms ultrabrief pulses (RUL) was preferred. For severe cases or situations in which RUL had been shown ineffective, patients underwent bilateral ECT with 1.0 ms brief pulses (BL). Empirical titration was performed in the first session for all patients (except if very urgent ones) and charge was calculated at 6× seizure threshold for RUL ECT and 2.5× the seizure threshold for BL ECT. Both acute and maintenance ECT were offered. A MECTA spECTrum 5000Q device (MECTA, Portland, OR, USA) was used and patients were monitored with 2 channels EEG and 1 channel ECG. Thiopental (1.5−2.5 mg/kg) was the anesthetic drug of choice although propofol (0.75−1.5 mg/kg) had to be used instead in a few sessions due to a thiopental shortage. Only those patients whose ECT sessions were already scheduled for that specific period were given propofol instead of thiopental. Succinylcholine (approximately 0.9 mg/kg) and atropine (0.3−0.6 mg) were also administered in every session.

As part of routine care, extensive data was recorded in every ECT session. Of interest for the present study, a chronometer was started at anesthetic induction and time elapsed was recorded when the electrical stimulus was delivered, when the patient ventilated spontaneously and when the patient became reoriented. Full reorientation was referred as to obtaining the same score on the exact 5 questions as before the ECT session: name, age, present location, month, year. A nurse frequently asked the patient regarding these questions, and when the pre and post-ECT scores were equal the time elapsed since anesthesia would be recorded.

Adverse effects immediately after every ECT session were recorded and the patient was also asked before every ECT session regarding side effects between the last and present session, including cognitive impairment and confusion. No post-ictal sedation was given to the patients to prevent agitation.

Study Design

This was a naturalistic, observational prospective study assessing the correlation between serum lithium levels and the time subjects would take to reorient themselves to person, time and space after the ECT procedure. Secondarily, we analyzed the correlation between both the motor and EEG seizure duration and the serum lithium levels in the same patients. Enrolled subjects would have blood samples drawn a few minutes before entering the ECT room where the procedure would take place.

Ethical Considerations and Statement of Interest

Due to its naturalistic observational design no changes of any sort were made in the patients’ treatment solely for the purpose of this study. Written informed consent regarding the study’s design, details and how their data, clinical notes, and charts would be handled was obtained from each patient. The study protocol has been reviewed and approved by the Institutional Review Board of the ethics committee of Cova da Beira University Healthcare Center (IRB approval no. 00012726).

Variables

Biochemical, ECT-related, and sociodemographic variables such as age and sex were collected. Serum lithium, TSH, T3, T4, Na＋, K＋, and Cl– levels were assessed via laboratorial analysis. Reorientation time was used as a surrogate marker for assessing the relationship between combined therapy with lithium plus ECT and the patients’ post-ictal cognitive impairment, as evidence suggests longer reorientation times to be associated with greater retrograde amnesia [22]. Time elapsed between anesthesia and the stimulus, motor seizure time, and EEG seizure time were measured using a digital stopwatch, as well as reorientation time. Sex, age, anesthetic drugs, ECT modality, number of titration sessions, and delivered stimulus charge were assessed directly through the subjects’ clinical files.

Statistical Analysis

Statistical analysis was performed with SPSS version 25.0 (IBM, Armonk, NY, USA). The effect of serum lithium on reorientation time was analyzed using Linear Mixed Effects Models with level 1 accounting for the ECT session and level 2 for the patients. Both the intercept and the slope were included as random effects in the model. A bivariate analysis was first run in order to assess the variables’ behavior in terms of slope and intercept for each patient before fitting a multilevel model. A step-up approach, which encompassed adding the variables according to their order of significance from the bivariate model and removing the ones which did not add significance to the model, was used in the multilevel analysis for finding the best fitting model. A sensibility analysis was then performed by omitting each subject at a time and running the analysis to ascertain if the end results would still be consistent and not due to the impact of an individual patient.

RESULTS

There was a total of 10 patients from the Cova da Beira Central University Hospital enrolled in the study which corresponded to a total of 114 ECT sessions. Twenty four of these entries were excluded due to incomplete data. Four of the remaining 90 reported a reorientation time > 1 hour (3,600 seconds) and were considered outliers, ending up being excluded, as in those specific cases patients actually fell asleep during the recovery. This rendered a total of 86 valid ECT sessions. Table 1 describes the sample’s demographic features and the number of ECT sessions each patient went through. Table 2 represents the sample’s distribution according to the selected clinical characteristics of the study. All patients except for one had the same electrode placement throughout their set of sessions. This patient took the first 3 sessions with a bilateral electrode placement with brief pulses and then changed to right unilateral with ultrabrief pulses. These results refer to patients whose serum lithium levels were mostly within 0.1 and 0.7 mmol/L (92%). Only 3 were first (titration) sessions.

Table 1

Age, sex, and number of sessions, by patient

Patient no.	Age	Sex	Number of sessions (%)
1	50	F	10 (11.6)
2	60	F	8 (9.3)
3	60	F	9 (10.5)
4	64	F	10 (11.6)
5	59	F	7 (8.1)
6	65	F	26 (30.2)
7	68	F	9 (10.5)
8	38	F	2 (2.3)
9	55	M	4 (4.7)
10	42	F	1 (1.2)
Total			86 (100)

Table 2

Distribution of the sample’s characteristics

Parameter				Number of session (%)
Electrode placement modality	Unilateral			35 (41)
	Bilateral			51 (59)
Anesthetic drug used	Thiopental			77 (89)
	Propofol			9 (11)
Titration session	No			83 (96)
	Yes			3 (4)
		Mean	SD		Minimum	Maximum
Age (yr)		56	10		38	68
Delivered charge (mC)		316	190		19	704
EEG seizure time (sec)		29	12		0	65
Motor seizure time (sec)		18	9		0	50
Serum lithium levels (mmol/L)		0.40	0.2		0.06	1.23
Time between anesthetic induction and stimulus (sec)		194	25		131	308
Time between the stimulus and patient’s reorientation (sec)		1,714	701		614	3,523
Serum TSH levels (uUI/ml)		3.1	2.5		0.9	21.6
Serum T3 levels (pg/ml)		2.6	0.6		1.4	3.9
Serum T4 levels (ng/dl)		0.9	0.2		0.7	1.5
Serum Na＋ levels (mmol/L)		140	4		112	148
Serum K＋ levels (mmol/L)		4.4	0.3		3.5	5.2
Serum Cl– levels (mmol/L)		103	4		82	110

SD, standard deviation; EEG, electroencephalogram; TSH, thyroid stimulating hormone; T3, triiodothyronine; T4, thyroxin; Na＋, sodium; K＋, potassium; Cl–, chlorine.

In a bivariate analysis serum lithium levels did not show an association with time to reorientation. Delivered charge showed a significant positive correlation with reorientation time (B = 1.2; p = 0.015). Regarding the type of anesthetic use, thiopental was associated with a significantly higher time to reorientation (B = 598; p = 0.007). No other variables showed significant associations with time to reorientation (Table 3).

Table 3

Bivariate analysis of the impact of ECT and patient factors on reorientation, motor seizure and EEG seizure times using an individual linear mixed model for each variable

Variable	Reorientation time				Motor seizure time				EEG seizure time
	Estimate	SE	pvalue		Estimate	SE	pvalue		Estimate	SE	pvalue
ECT session factors
Lithium	−226	372	> 0.05		−3.9	4	> 0.05		−5.1	7.2	> 0.05
ECT charge	1.2	0.5	0.015		−0.0002	0.009	> 0.05		0.006	0.015	> 0.05
RUL	−811	123	> 0.05		−1.2	3	> 0.05		0.9	5	> 0.05
BLa	-	-	-		-	-	-		-	-	-
Thiopental	598	220	0.007		3	2.6	> 0.05		−4	4	> 0.05
Propofola	-	-	-		-	-	-		-	-	-
Titration	123	311	> 0.05		0.7	4	> 0.05		−4	8	> 0.05
TSH	68	70	> 0.05		−0.9	0.9	> 0.05		0.1	1.5	> 0.05
T3	50	177	> 0.05		5.2	1.2	> 0.05		4.6	2.7	> 0.05
T4	−436	471	> 0.05		18	4	0.1		12	111	> 0.05
Na＋	21	16	> 0.05		0.1	0.3	> 0.05		0.2	0.4	> 0.05
K＋	15	222	> 0.05		−2	3	> 0.05		2.5	4	> 0.05
Cl–	−11	18	> 0.05		0.2	0.3	> 0.05		0.3	0.4	> 0.05
Patient factors
Age	17	16	> 0.05		−0.1	0.2	> 0.05		−0.4	0.4	> 0.05
Female	−357	321	> 0.05		−10	4	> 0.05		−1.4	6	> 0.05
Malea	-	-	-		-	-	-		-	-	-

ECT, electroconvulsive therapy; EEG, electroencephalogram; SE, standard error; Lithium, serum lithium levels; RUL, right unilateral ultrabrief pulse ECT; BL, bilateral brief pulse ECT; TSH, thyroid stimulating hormone; T3, triiodothyronine; T4, thyroxin; Na＋, sodium; K＋, potassium; Cl–, chlorine; -, not available.

aReference category.

Subsequently, a multivariate linear mixed model analysis was performed, using random slopes for each predictor, with time to reorientation as the outcome. A baseline model was obtained showing a calculated variance partition coefficient of 0.33, which means that patient individual differences accounted for about 33% of the variation in reorientation time. Patients were introduced as a random effect and ECT session variables were introduced as fixed effects using a step-up approach, and possible interactions between lithium and the used anesthetic drug as well as between lithium and electrode placement were ascertained. The best fitting model is represented in Table 4.

Table 4

Impact of serum lithium levels and other variables on re-orientation time

Parameter	Reorientation time
	Estimate	SE	pvalue
Serum lithium levels	−1,078	448	0.018
RUL	−814	157	< 0.001
BL (reference)	-	-	-
Thiopental	698	188	< 0.001
Propofol (reference)	-	-	-
Lithium*RUL interaction	1,447	576	0.014

Excluded variables: Age, sex, electroconvulsive therapy (ECT) charge, time elapsed between the anesthesia and the stimulus, titration session, thyroid stimulating hormone, triiodothyronine, thyroxine, sodium levels, potassium levels, and chlorine levels; Results shown according to the best fitting model obtained from a step-up approach using a multilevel analysis (Likelihood ratio chi-square = 29.8; df = 8; p < 0.001).

SE, standard error; RUL, right unilateral ultrabrief pulse ECT; BL, bilateral brief pulse ECT; Lithium*RUL interaction, interaction bet-ween serum lithium levels and RUL; *, interaction; -, not available.

We observed that lithium presented a significant negative relationship with the patients’ reorientation time only when its interaction with the ECT placement modality was considered in the model. The interaction between lithium and the ECT modality was found to play a significant role and to better examine this relation Figures 1 and 2 are presented, showing for RUL ECT no clear tendency regarding the effect of lithium levels on reorientation but, in contrast, for BL ECT a consistent decrease in reorientation time as lithium levels increased. For each 0.1 mmol/L of serum lithium increase, patients on BL ECT took on average 108 seconds (1'48'') less to recover. Figure 2 illustrates the representative model of the above mentioned.

Fig. 1

Correlation of serum lithium and reorientation time for each patient.

Fig. 2

Correlation of serum lithium and reorientation time for each pa-tient, split by electroconvulsive ther-apy (ECT) modality.

It was also found that patients submitted to RUL ECT took approximately 814 seconds (13'34'') less to reorientation than patients under BL ECT.

Regarding the anesthetic drug used, thiopental resulted in a recovery time of about 698 seconds (11'38'') longer when compared to propofol.

No significant model was obtained regarding motor seizure time and EEG seizure time.

A sensibility analysis evidenced that the final model was no longer statistically significant when subjects 6 and 7 were omitted.

Of note, no post-ictal acute confusional state or clinically evident level of cognitive impairment was reported after any of the presented ECT sessions, either by hospital personnel, the patient or caregivers.

DISCUSSION

The most striking finding of this study was to observe a negative correlation between serum lithium levels and the patients’ reorientation time, which only presented as statistically significant when considering its interaction with the ECT modality. Although such a finding may have been due to the inclusion of two specific subjects in the study, no significant positive correlation was found, contrary to what would be expected [18,19]. By assessing the serum lithium levels of patients undergoing BL ECT (and 1.0 millisecond brief pulses), we noticed that those with higher levels actually recovered faster from the procedure than those with lower levels. For those under RUL ECT (and 0.3 millisecond ultrabrief pulses) we found no influence of lithium in reorientation time. This relationship may be better perceived by analyzing the slopes of different patients when split by ECT type (Fig. 2). This observation contradicted our expectations. Previous authors have suggested either a positive correlation or no correlation at all [8,18,19,23], but there are no mentions either about a possible negative relationship between these two variables nor about a possible interaction between lithium and the ECT modality.

Despite the small amount of research studies mentioning this, there are some case reports that comment on the occurrence of adverse effects even in patients with subtherapeutic lithium levels [12,16,17,24], and el-Mallakh [25] has also expressed some theoretical concerns in his review regarding the fact that ECT may enhance central manifestations of lithium toxicity with therapeutic or even subtherapeutic levels, pointing out ECT as a plausible early trigger of lithium toxicity. In terms of physiology, a transient blood-brain barrier (BBB) breach caused by an intra-ictal hypertensive surge has been held responsible for the causative mechanism of some cognitive side effects such as amnesia due to the dispersion of circulating substances in the blood to the cerebral spinal fluid, though it is still unknown what substances are mostly implicated [26]. Normal diffusion of lithium into the brain occurs mainly based on the serum levels and also the length of exposure to the drug. In addition, lithium levels in the CSF are usually lower than in serum, with brain to serum ratios as low as 0.5 [27,28], although significant variation seems to occur along the 24-hour day cycle [28], with lower ratios in the morning. Finally, ECT-induced increase in BBB permeability may allow serum lithium to diffuse in larger amounts to the CSF, leading to the usual neurotoxic manifestations seen with higher serum lithium levels. However, this remains speculative and a study using animal models has not been able to demonstrate a statistically significant elevation of CSF lithium levels following ECT [12].

A key remark of the present study is its sample’s serum lithium levels, which were mostly within a low range, between 0.1 and 0.7 mmol/L, seemingly reflecting psychiatrists’ concerns with lithium plus ECT related side effects but at the same time maintaining patients on lithium therapy. The recommended therapeutic dosage of lithium for acute treatment aims at reaching serum levels between 0.8 and 1.2 mmol/L [2], although for depression lower levels might be indicated, or in maintenance therapy, where 0.4 to 1 mmol/L has been advocated to be sufficient, with a benign side effect profile [2,29]. So, in spite of the fact that most of our subjects had lithium levels in the lower range, the first hypothesis we could think of for possibly justifying such an observation would be of lithium not presenting a genuine linear relationship with the recovery time but actually a curved one. In fact, there is emerging evidence suggesting that lithium may even pose as a neuroprotective and neurotrophic drug in patients with bipolar disease, although not for every subtype of the disease [30]. This protective effect might not be exclusive for mood disorders, but having a more generalized effect, as lithium has also been studied for its potential in Alzheimer’s disease [31]. This could determine that low lithium levels could eventually be a protective factor against post-ECT cognitive impairment, and thus a negative correlation would be found until a certain threshold would be reached and patients would start exhibiting longer reorientation times as they would have higher lithium levels. However, this is also purely speculative and would require further investigation under a different study design.

The study also found that bilateral brief pulse ECT was associated on average with a longer reorientation time. Patients on bilateral brief ECT took about 13 minutes and 34 seconds longer than those on right unilateral ultrabrief ECT to reorient themselves. This observation is consistent with the literature [32], as bilateral brief pulse ECT has been associated with a higher proportion of adverse effects [1,23], even though it has been associated with better clinical outcomes in patients with depression [1].

Regarding the anesthetic drug used to induce sleep during the ECT procedure, thiopental was significantly associated with a longer recovery time, when compared to propofol, of about 11 minutes and 38 seconds. This observation is also very much consistent with the existing literature [33,34], and no significant interaction between serum lithium levels and the anesthetic drug was found.

In terms of age and sex, there was no clear relationship between subjects’ reorientation time, although evidence tells it otherwise. A study of Patel and associates has described a 12 times higher susceptibility for individuals above 60 years old to become acutely delirious and 6-fold higher odds of developing some sort of mild cognitive impairment following combined therapy of lithium and ECT when compared to those without lithium [19], although dosages and serum levels were not reported. It has also been described that people of geriatric age have a higher prevalence of post-ECT cognitive impairment [1] and that those who are older than 50 years old may have incremental risk of presenting signs of lithium toxicity [5].

Regarding EEG-assessed and motor seizure times, evidence points out that subjects taking lithium may be more prone to prolonged seizures, especially EEG-assessed when compared to observed motor seizures [18,35]. However, we did not find any significant relation between either observed motor or EEG-assessed seizure duration with the subjects’ serum lithium levels. EEG measuring is an operator dependent task which is determined on the basis of the clinician’s experience interpreting EEGs, and may leave this variable somewhat prompt to operator bias.

Also, no thyroid parameter or electrolytic ion was found to have a significant impact in the patients’ reorientation time, though there has been some evidence suggesting that the use of T3 in patients recovering from ECT may in fact improve their cognitive impairment, disregarding of their previous thyroid function. Interestingly enough, positive findings have shown that T3 may be used to improve the cognitive function of patients taking lithium, which has led some authors to propose the use of adjunctive treatment with thyroid hormones in order to minimize the negative cognitive effects of lithium and ECT [36].

The main shortcomings of this study were related to its structural design as it was subject to the many weaknesses inherent to naturalistic studies. The small sample size was a major limitation as well. The absence of statistically significant negative relationship between serum lithium levels and reorientation time in the sensibility analysis when two of the subjects were omitted suggests that either the number of sessions or their specific individual factors might have played a role in the results, although this must be interpreted cautiously as it may be due to the reduction of the statistical power that followed their omission from the analysis. Also, the absence of a control group, the variable number of sessions per subject, overrepresentation of females, were all limitations that might have influenced our results. Finally, it is important to mention that this study used the patients’ reorientation time as a surrogate marker for the relationship between combined therapy and cognition and that it was ascertained by variable elements of the nursing staff with different training using a 5 item only question list. No formal assessment of delirium after ECT as well as missed and abortive seizures were considered for the purpose of this study. This means that, since most of the literature, except for one study [18], addresses the above-mentioned clinical manifestations as their main outcome variables, our observations may not reflect the exact same clinical outcomes as most studies. However, reorientation time seems to be a good predictor of medium and long term cognitive function after ECT [22] and might even be a more relevant marker than a transient side effect such as post-ictal confusional states.

While this study adds more information for interpreting the combination of ECT with lithium it has also helped disclosing some of the uncertainty inherent to the field. Although the correlations found may be attributable to the impact of individual subjects in the study, the lack of a statistically significant positive relationship between serum lithium levels and reorientation time in patients with lithium levels on the lower therapeutic bound seems to be a robust finding in this study. It remains to be seen if these observations can be replicated in other studies and further research is still needed regarding different lithium levels under ECT with both right unilateral ultrabrief pulse and bilateral brief pulse ECT in order to bridge this chasm of the true impact lithium has in patients under ECT. As evidence is still scant we suggest that one should exercise caution whenever combining both therapies. However, it should be borne in mind that concomitant ECT and lithium might not necessarily need to be contraindicated and such recommendations may eventually present more harm than benefit in some patients. It might be plausible to consider maintaining lithium in some patients despite concomitant ECT, especially if bilateral (brief pulses) ECT. Performing a precise and tailor-made treatment for each patient is crucial, since lithium has a narrow therapeutic margin, meaning that in low to moderate levels it is therapeutic, but at slightly higher levels it may become neuro-toxic. Monitoring lithium levels along the ECT cycle should be a fundamental step, with target levels in the lower therapeutic bound (0.4−0.7 mmol/L), meaning that even during transient ECT induced BBB increase in permeability, there would be enough room before neurotoxic levels are reached.