ARTICLE

Severe Hyponatremia Is Often Drug

Induced: 10-Year Results of a Prospective
Pharmacovigilance Program
Elena Ramírez1,* , Amelia Rodríguez1, Javier Queiruga1, Irene García1, Lucía Díaz1, Lucía Martínez1,
Raúl Muñoz1, Mario Muñoz1, Hoi Y. Tong1 , José Carlos Martínez1, Alberto M. Borobia1 ,
Antonio J. Carcas1 and Jesús Frías1,*

We conducted a prospective evaluation of drug-induced severe hyponatremia  (adverse drug reaction (ADR)) through
the Prospective Pharmacovigilance Program from Laboratory Signals at Hospital over a period of 10 years. Cases
of serum sodium (Na(s)) < 116 mM were recorded from July 2007 to June 2017 (first period). Also cases of Na(s)
116–122 mM were recorded from July 2012 to June 2017 (second period). Drugs were the primary cause of severe
hyponatremia. The incidence rate of Na(s) < 116 mM by drugs was increased threefold over the decade. Compared
with other causes of hyponatremia, patients with adverse drug reaction–serum sodium (ADR-Na(s)) in the first
period were older (79 years (interquartile range (IQR) 68.6–89 vs. 65 years (IQR 48–81); P < 0.001) and were more
often women (70.8% vs. 48.9% men, P < 0.001); in the second period were also older (79 years (IQR 65.3–89) vs.
63 years (IQR 46–80.6); P < 0.001) and were more often women (70% vs. 53%, P = 0.002), and ADR-Na(s) occurred
more often in summer. The most frequent therapeutic groups of culprit drugs were the cardiovascular system and
nervous system. The 65.3% in the first period and 71.2% in the second period of the ADR-Na(s) cases responded to
hydration and had been diagnosed with hypovolemic hyponatremia.

Study Highlights

WHAT IS THE CURRENT KNOWLEDGE ON THE due to medication (adverse drug reaction–serum sodium
TOPIC? (ADR-Na(s))). The incidence of ADR-Na(s) is increasing in
 Hyponatremia is the most common electrolyte disturbance; the last decade and occurring more often during summer.
it is associated with fractures, falls, unsteadiness, dementia, and Seventy percent of ADR-Na(s) is hypovolemic hyponatremia
greater mortality risk. Hyponatremia has multiple etiologies, and 30% is a syndrome of inappropriate antidiuretic hormone
and distinguishing the cause is difficult. secretion (SIADH). Low urea (SIADH) and low chlorine
WHAT QUESTION DID THIS STUDY ADDRESS? (hypovolemic) concentrations allow distinguishing between
 Assess the causes of severe hyponatremia. Predictive factors them.
and potential preventability of drug-induced severe hypona- HOW MIGHT THIS CHANGE CLINICAL PHARMA-
tremia were assessed. COLOGY OR TRANSLATIONAL SCIENCE?
WHAT DOES THIS STUDY ADD TO OUR KNOW-  This is an important finding that suggests clinical awareness
LEDGE? should be increased, especially in older female patients using
 In this manuscript it is described how hyponatremia cau- cardiovascular drugs or nervous system drugs. Changes in drug
sality can be evaluated and that severe hyponatremia is often prescriptions are essential to avoid this serious ADR.

Hyponatremia is defined by low serum sodium (Na(s)) concentra- hyponatremia at some point during their hospitalization,3 and
tion and is the most common electrolyte disturbance in the com- hyponatremia has been observed in 18% of aged care facility
munity.1 Some 6% of people over 55 years old have hyponatremia; residents.4 Na(s) concentration is determined by the sum of the
this percentage increases to 11.6% in people older than 75 years.1 exchangeable (osmotically active) portions of the body’s sodium
During the summer, hyponatremia prevalence increases over 33% and potassium divided by total body water (Edelman equation;
in elderly patients, older than 65 years, with a significant correla- Na(s) = (Nae + Ke)/TBW).5 Hyponatremia occurs as a by-product
tion between outdoor temperature and hyponatremia prevalence.2 of unmatching the intake of sodium, potassium, and water with
Over 40% of elderly patients admitted to intensive care experience the corresponding losses.6 Chronic hyponatremia (>48 hours in

1
Department of Clinical Pharmacology, La Paz University Hospital-Carlos III, IdiPAZ, School of Medicine, Autonomous University of Madrid, Madrid,
Spain. *Correspondence: Elena Ramírez (elena.ramirezg@uam.es); and Jesús Frías (jesus.frias@uam.es)
Received December 27, 2018; accepted June 8, 2019. doi:10.1002/cpt.1562

1362 CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 106 NUMBER 6 | DECEMBER 2019
 ARTICLE

duration) is associated with fractures, falls, unsteadiness, atten- hospital acquired or community acquired, through a PPLSH in
tion deficits, and greater mortality risk.7–9 Acute hyponatremia hospitalized patients over a period of 10 years. Based on these data,
with the same magnitude as chronic hyponatremia is likelier to we aimed to estimate the incidence, causes, seasonality, clinical pre-
induce more drastic symptoms, such as seizures or coma, or clin- sentation, outcomes, and potential preventability of this particular
ical manifestation of brain edema than can lead to death due to ADR.
cerebral herniation.10 In addition, brain complications can appear
following the treatment of hyponatremia. Central pontine myeli- RESULTS
nolysis is the loss of myelin in the pons that occurs in patients who Over the first 5 years of the study, the number of hospital admis-
had their hyponatremia corrected too rapidly.10 Improvements sions was 238,311 (from July 2007 to June 2012). There were
in Na(s) concentration in patients with hyponatremia have been 1,881,531 Na(s) level measurements; of these, 376 cases of Na(s)
shown to be associated with a reduction in overall mortality.11 It < 116 mM were detected, and of them, 72 cases in 69 patients were
is also seems likely that hyponatremia itself leads to cognitive im- categorized as ADRs (Figure  1a). During the second 5  years of
pairment and additionally, that correction of hyponatremia can the study, the number of hospital admissions was 225,862 (from
lead to improved cognition.12 July 2012 to June of 2017). There were 2,048,447 Na(s) level mea-
Clinical management of hyponatremia is based on treating the surements; of these, 906 cases of Na(s) < 122 mM were detected,
underlying causes; however, an accurate determination of the type and of these, 333 cases in 318 patients were categorized as ADRs
of hyponatremia is a challenge, particularly in elderly patients.13 (Figure 1b). The incidence rate of drug-induced very severe hypo-
It is relatively easy to exclude hypervolemic hyponatremia due to natremia (Na(s) < 116 mM) per 10,000 patients over the 10 years
heart failure, chronic kidney disease, and chronic liver disease, of the study was 3.47 (95% confidence interval (CI), 1.08–8.77),
based on the patient’s history and physical examination. However, increasing from 2.58 (95% CI, 0.62–7.22) in 2007 to 9.57 (95%
the differential diagnosis between a syndrome of inappropriate CI, 4.80–17.08) in 2017. The incidence rate of drug-induced se-
antidiuretic hormone secretion (SIADH) and hypovolemic hy- vere hyponatremia (Na(s) 116–122 mM) per 10,000 patients the
ponatremia is difficult because there is currently no reliable bio- last 5 years of the study was 10.80 (95% CI, 5.49–18.39), increas-
marker of hydration in older people and clinical examinations are ing from 7.95 in 2012 (95% CI, 3.45–14.42) to 12.38 in 2017
unreliable.14 In addition, knowing the type of hyponatremia is (95% CI, 6.92–20.96) (Figure  2a). Table  1 shows the numbers
essential, not only for appropriate clinical management but also of cases, percentages, incidence rates, and confidence intervals of
to prevent the rechallenging of a drug-induced hyponatremia.15 incidence rates corresponding to the causes associated with hypo-
In the Hyponatremia International Registry, the common initial natremia at admission or during hospitalization. Compared with
clinical management of hyponatremia was fluid restriction (35%), other causes of hyponatremia, patients with ADRs in the first pe-
followed by administration of isotonic (15%) or hypertonic saline riod were older (79 years (interquartile range (IQR) 68.6–89) vs.
(2%), and tolvaptan (5%).16 Additionally, the majority of patients 65  years (IQR 48–81); P  <  0.001) and were more often women
underwent withdrawal of an agent that would interfere with water (70.8% vs. 48.9%, P < 0.001); the patients with ADRs in the sec-
excretion.17 ond period were also older (79 years (IQR 65.3–89) vs. 63 years
In older patients, inappropriate prescribing is highly prevalent (IQR 46–80.6); P  <  0.001) and were also more often women
and is associated with an increased risk of morbidity, mortality, and (70% vs. 53%, P = 0.002). The distribution of hyponatremia by
healthcare utilization.18 Inappropriate prescribing can be detected periods per month is shown in Figure 2b. Age (odds ratio (OR)
using explicit or implicit prescribing indicators. The Screening 1.028; 95% CI, 1.021–1.036; P < 0.001), female sex (OR 1.990;
Tool of Older People’s Prescriptions (STOPP) criteria were first 95% CI, 1.512–2.643; P < 0.001), and the months from June to
published in 2008.19 Due to expanding therapeutics, the criteria September were significantly associated with ADR-Na(s) (June
were updated in 2015.20 For example, in STOPP criteria it is rec- OR 3.731, 95% CI, 1.425–11.111, P = 0.005; July OR 4.310, 95%
ommended not to use thiazide diuretics or selective serotonin re- CI, 1.524–13.514, P = 0.004; August OR 6.623, 95% CI, 2.309–
uptake inhibitors in patients with Na(s) < 130 mM. 21.277, P = 0.001; September OR 4.348, 95% CI, 1.558–10.101,
In recent decades, the availability of large, computerized clinical P  =  0.003). Maximum temperature was significantly associated
databases linked to electronic medical records (EMRs) has been with ADR-Na(s) (OR 1.076, 95% CI, 1.033–1.132, P =  0.008).
useful in implementing prospective programs for the detection The logistic model, ROC (receiver operating characteristic) curve,
of adverse drug reactions (ADRs) and to aid clinicians in react- ADR ~ age + sex + month + Tmax + Na(s) level is represented in
ing quickly and appropriately.21 We have developed a prospective Figure  2b. A cross-validation, 1,000 simulations of the model,
pharmacovigilance program based on the systematic detection of was performed: Sets A (32.5%), B (32.3%), and C (32.2%) cor-
predefined abnormal laboratory signals, through the laboratory rectly classified 70.1%, 69%, and 70.1% of cases, respectively.
information system at our hospital (Pharmacovigilance Program
from Laboratory Signals at Hospital, PPLSH). Screening for spe-
cific anomalous laboratory data allows us to monitor a large num- Culprit drugs
ber of patients with limited resources, thus accessing high-quality There were 151 culprit drugs found during the first period (hypona-
ADR information in a timely manner.22 The aim of this study tremia < 116 mM; 72 cases in 69 patients) and 635 found during the
was to detect and report on drug-induced severe hyponatremia second period (hyponatremia < 122 mM; 333 cases in 318 patients).
(adverse drug reaction–serum sodium (ADR-Na(s))), whether The median (range) of suspected drugs per case was 2 (1–5). The

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 106 NUMBER 6 | DECEMBER 2019 1363
ARTICLE

Figure 1  Methodology and flowchart of signal-sodium from the Pharmacovigilance Program From Laboratory Signals at Hospital. (a) First
period: Query to Laboratory Database from July 1, 2007 to June 30, 2012. (b) Second period: Query to Laboratory Database from July 1, 2012
to June 30, 2017. ADR, adverse drug reaction. [Colour figure can be viewed at wileyonlinelibrary.com]

most frequent therapeutic groups were cardiovascular system drugs frequency of neurological symptoms such as coma or convulsions
(50.3% vs. 60.8%) and nervous system drugs (32.5% vs. 19.7%). The was 17.8% (72/405), and neurological symptoms were the cause of
most frequent drug during both periods was hydrochlorothiazide death in 8% of the cases. The median (range) latency in days was 41
(15.3% vs. 18.3%). Table  2 shows the drugs causing 405 cases of (1–3650) days. The median (range) of recovery days was 8 (2–112)
severe hyponatremia. A positive rechallenge was recorded in 17 pa- days. Five cases of pontine myelinolysis occurred in the very severe
tients: 8 patients with hydrochlorothiazide, 4 cases with spirono- cases group. Characteristics of the patients are listed in Table 3.
lactone, 2 cases with furosemide, 1 case with fentanyl or Aesculus
hippocastanum or domperidone, and 1 patient had two positive Types of hyponatremia
rechallenges with quetiapine. The median (range) to rehospitaliza- The 65.3% in first period and 71.2% in second period of cases
tion due to a positive rechallenge was 1 month (15 days–8 years). responded to isotonic saline and were diagnosed as hypovole-
mic hyponatremia regardless of their apparent hydration status.
Patient characteristics Hypovolemic hyponatremia was divided into thiazide and thia-
Out of 387 patients with drug-induced severe or very severe hypo- zide-like induced hyponatremia (TIH) and non–thiazide-induced
natremia, 381 were adults. Of these, 73.6% were older than 65 years hypovolemic hyponatremia. In the rest of cases, SIADH was diag-
of age. The 88.6% of cases resulted in hospitalization, and only nosed. Out of SIADH cases, 103 (57.9%) responded to fluid re-
44 (10.9%) cases occurred during hospitalization. Six cases (1.5%) striction and oral sodium chloride, 69 (38.8%) to hypertonic saline,
of hyponatremia in children occurred during hospitalization. and 6 (3.3%) to tolvaptan (available in the hospital from 2010). In
The most frequent underlying condition was cardiovascular dis- comparison with hypovolemic hyponatremia, SIADH cases had
ease (52.5%), followed by dementia (31.3%) and diabetes mellitus a significantly lower urea concentration (48  mg/dL vs. 17.1  mg/
(29.2%). The most frequent symptom was falls with serious con- dL, P < 0.001). Patients with hypovolemic hyponatremia had de-
sequences (fractures or brain damage), in 58.57% of the cases. The creased creatinine clearance compared with SIADH (60.3  mL/

1364 VOLUME 106 NUMBER 6 | DECEMBER 2019 | www.cpt-journal.com

 ARTICLE

Figure 2  (a) Incidence rate per 10,000 patients of hyponatremia, ADR and non-ADR. Above is represented hyponatremia < 116 mM over
a 10-year period, below is represented hyponatremia between 116–122 mM of the last 5-year period. (b) The distribution of hyponatremia
by periods per month. Receiver operating characteristic (ROC) curve to evaluate the seasonality, age, sex, and serum sodium level of ADR
patients compared with non-ADR. ADR, adverse drug reaction; AUC, area under the time concentration curve; CI, confidence interval;
Tmax, time to maximum concentration. [Colour figure can be viewed at wileyonlinelibrary.com]

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 106 NUMBER 6 | DECEMBER 2019 1365
ARTICLE

Table 1  Breakdown by diagnosis of severe hyponatremia recorded at La Paz University Hospital by periods: first period (from
July 2007 to June 2012), second period (from July 2012 to June 2017)
Poisson 95% con-
Incidence ratea  fidence interval of
N of % of (per 10,000 incidence rate (per
Signal category Causes cases cases patients) 10,000 patients) P value
First period Hemorrhage or masses CNS 76 20.2 3.19 1.08–8.77 0.860
Hyponatremia  Neoplasia 74 19.7 3.11 1.08–8.77 1.000
< 116 mM
Drugs 72 19.2 3.02 1.08–8.77 ref.
Heart failure or cirrhosis 38 10.1 1.59 0.24–5.57 0.071
Pulmonary infection 27 7.2 1.13 0.24–5.57 0.012
Gastrointestinal losses 25 6.7 1.05 0.24–5.57 0.005
Advanced renal failure 21 5.6 0.88 0.03–3.69 0.002
Endocrine disorders 15 4.0 0.63 0.03–3.69 <0.001
Primary polydipsia 9 2.4 0.38 0.03–3.69 <0.001
Pseudohyponatremia 7 1.9 0.29 0.03–3.69 <0.001
Low dietary solute intake  2 0.5 0.08 0.03–3.69 <0.001
Exercise-associated 2 0.5 0.08 0.03–3.69 <0.001
Other 8 2.1 0.34 0.03–3.69 <0.001
Total 376 100.0 15.78 10.67–27.22  
Second period Drugs 333 36.8 14.74 8.40–23.49 ref.
Hyponatremia  Heart failure or cirrhosis 111 12.3 4.91 1.63–10.24 <0.001
< 122 mM
Gastrointestinal losses 93 10.3 4.12 1.63–10.24 <0.001
Advanced renal failure 79 8.7 3.50 1.09–8.77 <0.001
Neoplasias 71 7.8 3.14 1.09–8.77 <0.001
Hemorrhage, CNS masses 65 7.2 2.88 0.62–7.23 <0.001
Endocrine disorders 36 4.0 1.59 0.24–5.57 <0.001
Pulmonary infection 34 3.8 1.51 0.24–5.57 <0.001
Primary polydipsia 25 2.8 1.11 0.24–5.57 <0.001
Low dietary solute intake 14 1.6 0.62 0.03–3.69 <0.001
Exercise-associated 12 1.3 0.53 0.03–3.69 <0.001
Pseudohyponatremia 10 1.1 0.44 0.03–3.69 <0.001
Other 23 2.5 1.02 0.24–5.57 <0.001
Total 906 100.0 40.11 29.42–54.47 <0.001
Values of drug-induced hyponatremia are bold.
CNS, central nervous system; ref., reference.
a
Excluding 129 and 115 errors (preanalytical errors) in hyponatremia < 116 mM and hyponatremia < 122 mM, respectively.

minute/1.73 m2 vs. 93.7 mL/minute/1.73 m2 , P < 0.001). In com- (10,000 iterations) was 0.991 (95% CI, 0.990–0.992). Figure 3
parison with SIADH, patients with TIH had reduced chlorine shows the details of the model.
(101.4 mg/dL vs. 80.2 mg/dL, P < 0.001). Table 3 shows the dif-
ferences between the types of hyponatremia by periods. Preventability
The STOPP showed that out of 405 cases of drug-induced hy-
Predictive factors ponatremia, 305 (75.3%) prescriptions were made to patients
The inverse of Na(s) (dependent variable) was related to age, older than 65  years, of which, 66 (23.3%) drugs—23 (46%) in
sex (female), type of hyponatremia (hypovolemic or normovole- the first period and 43 (12.9%) in the second period—met some
mic), underlying conditions, origin (community or hospital STOPP criteria; of these 66 drugs, 45 (68.2%) met one crite-
acquired), manifestations of hyponatremia (falls, weakness or rion, 15 (22.7%) met two criteria, and 6 (9.1%) met three criteria.
deterioration, fatigue, confusion, neurological symptoms, un- Cardiovascular system (Table 4, section B) drugs were the group
consciousness, seizures), latency, number of related drugs, or that met the most STOPP criteria (28; 42.4%), followed by drugs
maximum or minimum outdoor temperature (independent that predictably increase the risk of falls in older people (Table 4,
variables). The adjusted coefficient of determination (R 2) of section K) (14; 21.2%) and central nervous system and psychotro-
the linear regression model with bootstrap robust estimation pic drugs (Table 4, section D) (9; 13.6%).

1366 VOLUME 106 NUMBER 6 | DECEMBER 2019 | www.cpt-journal.com

 Table 2  Drug groups and subgroups causing 405 cases of drug-induced severe hyponatremia by periods: first period, from July 2007 to June 2012, 151
culprit drugs in the drug-induced very severe hyponatremia (<116 mM) category; second period, from July 2012 to Jun 2017, 635 in the drug-induced severe
hyponatremia (<122 mM) category
Number Number (%) Number (%)
Signal category of drugs Drugs (group) of drugs Drugs (subgroup) of drugs Drugs
First period 151 Nervous system 49 (32.5) Antidepressants 24 (15.9) Mirtazapine (6), sertraline (4), escitalopram (3),
Hyponatremia  duloxetine (2), citalopram (2), venlafaxine (2),
< 116 mM paroxetine (2), trazodone (2), fluoxetine (2)
Antipsychotics 8 (5.3) Risperidone (4), levomepromazine (2), haloperidol (1),
quetiapine (1)
Antiepileptics 8 (5.3) Valproic acid (2), pregabalin (2), carbamazepine (2),
gabapentin (1), levetiracetam (1)
Anxiolytics 4 (2.7) Lorazepam (2), alprazolam (1), lormetazepam (1)
Hypnotics and sedatives 2 (1.3) Midazolam (1), zolpidem (1)
Opioids 2 (1.3) Morphine (1), fentanyl (1)
Anticholinergic agents 1 (0.7) Biperiden (1)
a,b,c
Cardiovascular 76 (50.3) Thiazides 23 (15.2) Hydrochlorothiazide (23)
system
Angiotensin II antagonistsa 14 (9.3) Valsartan (5), telmisartan (4), losartan (2), candesartan
(2), olmesartan (1)
Low-ceiling diuretics, excluding thiazides 8 (5.3) Chlortalidone (4), indapamide (4)
High-ceiling diuretics 8 (5.3) Furosemide (8)
b
ACE inhibitors. plain 8 (5.3) Enalapril (8)

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 106 NUMBER 6 | DECEMBER 2019

Potassium-sparing agentsc 7 (4.6) Amiloride (6), spironolactone (1)
Selective calcium channel blockers with 3 (2.0) Amlodipine (2), nifedipine (1)
mainly vascular effects
Beta blocking agents 2 (1.3) Atenolol (2)
Class 1 and 3 antiarrhythmic drugs 1 (0.7) Amiodarone (1)
Capillary stabilizing agents 1 (0.7) Diosmin (1)
Selective calcium channel blockers with 1 (0.7) Verapamil (1)
direct cardiac effects
Antibiotics for sys- 5 (3.3) Macrolides 1 (0.7) Clarithromycin (1)
temic use Glycopeptides 1 (0.7) Vancomycin (1)
Sulfonamides and trimethoprim 1 (0.7) Trimethoprim sulfamethoxazole (1)
Other aminoglycosides 1 (0.7) Tobramycin (1)
Beta-lactam antibacterials. Penicillins 1 (0.7) Penicillin combinations (1)
Alimentary tract and 4 (2.7) Drugs for constipation 2 (1.3) Lactulose (1), sodium phosphate (1)
metabolism Blood glucose lowering drugs, except 1 (0.7) Clorpropamide (1)
insulins
Drugs for peptic ulcer and gastro-esoph- 1 (0.7) Omeprazole (1)
ageal reflux disease

1367
ARTICLE

(Continues)
 Table 2  (Continued)

1368
Number Number (%) Number (%)
Signal category of drugs Drugs (group) of drugs Drugs (subgroup) of drugs Drugs
ARTICLE

Antineoplastic and 6 (4.0) Other antineoplastic agents 4 (2.7) Cisplatin (3), oxaliplatin (1)
immunomodulating Antimetabolites 2 (1.3) Fluorouracil (1), capecitabine (1)
agents
Various 2 (1.3) Diagnostic agents 2 (1.3) Iobitridol (1), iodinated contrast (1)
Herbology products 2 (1.3) Horse tail 1 (0.7)  
Horse chestnut 1 (0.7)  
Systemic hormonal 2 (1.3) Posterior pituitary lobe hormones 1 (0.7) Desmopressin (1)
preparations, exclud- Corticosteroids for systemic use 1 (0.7) Methylprednisolone (1)
ing sex hormones
and insulins
Musculoskeletal 1 (0.7) Anti-inflammatory and antirheumatic 1 (0.7) Dexketoprofen (1)
system products. Nonsteroids
Respiratory system 1 (0.7) Cough suppressants 1 (0.7) Dextromethorphan (1)
     
Genito-urinary 1 (0.7) Urologic drugs 1 (0.7) Tolterodine (1)
system and sex
hormones
Other 2 (1.3) Sodium polystyrene sulfonate 1 (0.7)  
Clinical trial drugs 1 (0.7) Dalcetrapib (1)
Second period 635 Nervous system 125 Antidepressants 49 (7.72) Venlafaxine (9), escitalopram (8), mirtazapine (7),
Hyponatremia  (19.69) trazodone (7), duloxetine (6), sertraline (4), paroxetine (3),
< 122 mM amitriptyline (3), mianserin (1), citalopram (1)
Antiepileptics 27 (4.25) Carbamazepine (9), valproic acid (4), gabapentin (3),
oxcarbazepine (2), phenytoin (2), lacosamide (2),
lamotrigine (2), eslicarbazepine (1), levetiracetam (1),
zonisamide (1)
Antipsychotics 25 (3.94) Quetiapine (7), olanzapine (3), risperidone (3), amisulpride
(2),haloperidol (2), tiapride (2), levomepromazine (1),
clotiapine (1), zuclopenthixol(1), fluphenazine (1),
clozapine (1), levosulpiride (1)
Opioids 10 (1.57) Fentanyl (4), morphine (3), tapentadol (2), buprenorphine
(1)
Antidementia drugs 7 (1.10) Memantine (3), galantamine (2), rivastigmine (2)
Anxiolytics 4 (0.63) Lorazepam (2), alprazolam (1), diazepam (1)
Hypnotics and sedatives 1 (0.16) Zolpidem (1)
Dopaminergic agents 1 (0.16) Pramipexol (1)
Parasympathetic drugs 1 (0.16)  
c,d,e
Cardiovascular 386 Thiazides 116.00 Hydrochlorothiazide (116)
system (60.79) (18.27)

VOLUME 106 NUMBER 6 | DECEMBER 2019 | www.cpt-journal.com

(Continues)
 Table 2  (Continued)
Number Number (%) Number (%)
Signal category of drugs Drugs (group) of drugs Drugs (subgroup) of drugs Drugs
Potassium-sparing agents 73 (11.50) Spironolactone (44), amiloride (23), eplerenone (6)
High-ceiling diuretics 64 (10.08) Furosemide (62), torasemide (2)
Angiotensin II antagonists 42 (6.61) Valsartan (17), losartan (9), olmesartan (6), telmisartan
(4), irbesartan (4), candesartan (1), eprosartan (1)
Low-ceiling diuretics, excluding thiazides 28 (4.41) Chlortalidone (15), indapamide (13)
ACE inhibitors, plain 27 (4.25) Enalapril (22), perindopril (2), lisinopril (2), ramipril (1)
Selective calcium channel blockers with 23 (3.62) Amlodipine (19), manidipine (2), nifedipine (1), felodipine
mainly vascular effects (1)
Beta blocking agents 4 (0.63) Atenolol (2), bisoprolol (1), propranolol (1)
Class 1 and 3 antiarrhythmic drugs 3 (0.47) Amiodarone (3)
Other cardiac preparations 2 (0.31) Ranolazine (1), ivabradine (1)
Selective calcium channel blockers with 1 (0.16) Diltiazem (1)
direct cardiac effects
Cardiac glycosides list 1 (0.16) Digoxin (1)
Cardiac stimulants excluding cardiac 1 (0.16) Dopamine (1)
glycosides
Arteriolar smooth muscle, agents acting on 1 (0.16) Hydralazine (1)
Antineoplastic and 39 (6.14) Other antineoplastic agents 18 (2.83) Cisplatin (7), oxaliplatin (3), bevacizumab (1), sunitinib (1),

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 106 NUMBER 6 | DECEMBER 2019

immunomodulating trastuzumab (1), dasatinib (1), irinotecan (1), cetuximab
agents (1), lapatinib (1), bexarotene (1)
Antimetabolites 9 (1.42) Capecitabin (4), methotrexate (2), fluorouracil (1),
pemetrexed (1)
Plant alkaloids and other natural 4 (0.63) Docetaxel (2), paclitaxel (1), vincristine (1)
products
Immunosuppressant drugs 3 (0.47) Adalimumab (1), cyclosporine (1), azathioprine (1)
Alkylating agents 2 (0.31) Cyclophosphamide (2)
Cytotoxic antibiotics and related 2 (0.31) Doxorubicin (2)
substances
Clinical trial 1 (0.16) Bococizumab (1)
Alimentary tract and 19 (2.99) Drugs for peptic ulcers and 6 (0.94) Omeprazole (4), pantoprazole (1), esomeprazole (1)
metabolism gastroesophageal reflux disease
Propulsives 6 (0.94) Domperidone (2), metroclopramide (2), cinitapride (1),
clebopride (1)
Drugs for constipation 5 (0.79) Bisacodyl (1), lactitol (1), macrogol (1), lactulose (1), lauryl
sulfate (1)
Antacid drugs (stomach acid drugs) 1 (0.16) Magaldrate (1)
Drugs for functional bowel disorders 1 (0.16) Mebeverine (1)

1369
ARTICLE

(Continues)
 Table 2  (Continued)

1370
Number Number (%) Number (%)
Signal category of drugs Drugs (group) of drugs Drugs (subgroup) of drugs Drugs
ARTICLE

Antibiotics for sys- 14 (2.20) Other beta-lactam antibacterials 6 (0.94) Ceftriaxone (2), meropenem (1), cefotaxime (1), cefepime
temic use (1), cefuroxime (1)
Beta lactam antibiotics, penicillins 3 (0.47) Penicillin combinations (2), cloxacillin (1)
Other antibacterial drugs 2 (0.31) Linezolid (2)
Quinolone antibacterials 1 (0.16) Ciprofloxacin (1)
Glycopeptide antibacterials 1 (0.16) Vancomycin (1)
Antimycotics for systemic use 1 (0.16) Amphotericin B (1)
Respiratory system 17 (2.68) Adrenergics, inhalants 10 (1.57) FDC of formoterol (3), salbutamol (3), salmeterol (3) and
fluticasone
Antihistamines for systemic use 3 (0.47) Desloratadine (1), bilastine (1), fexofenadine (1)
Decongestants and other nasal prepara- 2 (0.31) Fluticasone (2)
tions for topical use
Other drugs for obstructive airway dis- 2 (0.31) Budesonide (2)
eases, inhalants
Herbology products 9 (1.42) Horse tail 3 (0.47)  
Plantago ovata forsk 2 (0.31)  
Senna leaves 1 (0.16)  
Alpine tila 1 (0.16)  
Unknown herbal products 2 (0.31)  
Musculoskeletal 6 (0.94) Anti-inflammatory and antirheumatic 5 (0.79) Dexketoprofen (3), rofecoxib (1), diclofenac (1)
system products, nonsteroids
Muscle relaxants, centrally acting agents 1 (0.16) Baclofen (1)
Systemic hormonal 5 (0.79) Corticosteroids for systemic use, plain 3 (0.47) Prednisone (2), methylprednisolone (1)
preparations, exclud- Posterior pituitary lobe hormones 1 (0.16) Desmopressin (1)
ing sex hormones
and insulins Parathyroid hormones and analogues 1 (0.16) Teriparatide (1)
Blood and blood 4 (0.63) Antithrombotic agents 4 (0.63) Bemiparin (1), acenocoumarol (2), acetylsalicylic acid (1)
forming organs
Genito-urinary 3 (0.47) Urologicals 3 (0.47) Solifenacin (3)
system and sex
hormones
Sensory organs 4 (0.63) Antiglaucoma preparations and miotics 4 (0.63) Acetazolamide (4)
Various 3 (0.47) Diagnostic agents 3 (0.47) Sodium amidotrizoate (2), iobitridol (1)
Other 1 (0.16) Human immunoglobulins 1 (0.16)  
ACE, angiotensin converting enzyme; FDC, fixed dose combination.
a
Nine cases of fixed combination thiazide and angiotensin II antagonists. bFive cases of fixed combination thiazide and ACE inhibitors. cFour cases of fixed combination thiazides and potassium-sparing agents.
d
Twenty-three cases of fixed combination thiazides and potassium-sparing agents. eTwenty-three cases of fixed combination thiazide and angiotensin II antagonists.

VOLUME 106 NUMBER 6 | DECEMBER 2019 | www.cpt-journal.com

 Table 3  Comparison of characteristics of patients with drug-induced very severe hyponatremia (Na(s) < 116 mM) and drug-induced severe hyponatremia (Na(s)
116–122 mM) by periods: first period (from July 2007 to June 2012), second period (from July 2012 to June 2017)
Second period First period
Total very
Total ADR severe Total severe Very severe Very severe

Na (s) Na (s)
  < 122 mM Na < 116 mM 116–122 mM Na < 116 mM Na < 116 mM P value    
Average serum sodium concentration, mM (SD) 117.2 (4.3) 112.9 (3.6) 120.0 (1.7) 112.6 (4.0) 113.0 (4.2) <0.001  
No. of cases (%) 405 161 (39.8) 244 (60.2) 89 (55.3) 72 (44.7) 0.002  

No. of patients (%) 387 156 (40.3) 231 (59.7) 87 (55.8) 69 (44.2) 0.007
No. of children (<18 years) (%) 6 (1.5) 2 (1.3) 4 (1.7) 1 (1.2) 1 (1.4) 1.0
No. of adults (>18 years) (%) 381 (98.5) 154 (98.7) 227 (98.3) 85 (97.7) 68 (98.6) 1.0
 0–1 years (%) 2 (0.5) 0 2 (0.9) 0 0 1.0
 2–5 years (%) 2 (0.5) 2 (1.3) 0 2 (2.3) 0 1.0
 6–11 years (%) 1 (0.3) 0 1 (0.4) 0 0 1.0
 12–17 years (%) 1 (0.3) 0 1 (0.4) 0 1 (1.4) 1.0
 Young adults (ages 18–45 years) (%) 16 (4.1) 8 (5.1) 8 (3.5) 5 (5.7) 3 (4.3) 0.471
 Middle-aged adults (age 46–64 years) (%) 80 (20.7) 32 (20.5) 48 (20.8) 18 (20.7) 14 (20.3) 1.0
 Older adults (age > 65 years) (%) 285 (73.6) 113 (73.0) 171 (74.0) 63 (72.4) 50 (72.5) 0.876
 Octogenarian (age > 80 ages) (%) 184 (47.5) 76 (48.7) 110 (47.5) 42 (48.3) 34 (49.3) 0.894
Origin

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 106 NUMBER 6 | DECEMBER 2019

 Community acquired (%) 359 (88.6) 147 (91.3) 214 (87.7) 81 (91.0) 66 (91.7) 0.372  
a  
 Hospital acquired (%) 44 (10.9) 14 (8.7) 30 (12.3) 8 (9.0) 6 (8.3) 0.351
Adult population
 Median age, years (range) 79 (18–100) 79 (27–99) 74 (18–100) 79 (28–99) 79 (27–98) 0.576  
 Female sex (%) 285 (70.7) 118 (73.3) 167 (68.4) 66 (74.2) 51 (70.8) 0.440  

Underlying condition
 Cardiovascular diseases (%) 203 (52.5) 77 (49.4) 124 (53.7) 44 (49.4) 33 (45.8) 1.0  
 Dementia (%) 131 (33.9) 78 (50.0) 53 (22.9) 44 (49.4) 34 (47.2) <0.001  

 Diabetes mellitus (%) 113 (29.2) 43 (27.6) 70 (30.3) 24 (27.0) 19 (26.4) 0.760
 Neoplasia (%) 44 (11.4) 17 (10.9) 27 (10.4) 9 (10.1) 6 (8.3) 0.820
 Hypersplenism (%) 28 (7.2) 10 (6.4) 18 (7.8) 6 (6.7) 4 (5.6) 0.773
 Hypothyroidism 16 (4.1) 6 (3.8) 10 (4.3) 3 (3.4) 3 (4.2) 0.799
 Diarrhea or dehydration (%) 18 (4.7) 8 (5.1) 10 (4.3) 5 (5.6) 3 (4.2) 0.736
 Moderate renal impairment (eGFR 30–60 mL/minute) 16 (4.1) 3 (1.9) 10 (4.3) 1 (1.1) 2 (2.8) 1.0

(Continues)

1371
ARTICLE
 Table 3  (Continued)

1372
Second period First period
Total very
Total ADR severe Total severe Very severe Very severe
ARTICLE

Na (s) Na (s)
  < 122 mM Na < 116 mM 116–122 mM Na < 116 mM Na < 116 mM P value    
Main symptom
 Falls (%) 140 (34.6) 45 (28.0) 94 (38.5) 25 (28.1) 20 (27.8) 0.132  
 Fractures or brain injury (%) 82 (20.2) 15 (9.3) 67 (27.5) 8 (9.0) 7 (9.7) <0.001  

 Weakness or deterioration (%) 90 (22.2) 47 (29.2) 45 (18.4) 26 (29.2) 21 (29.2) 0.067

 Fatigue (%) 56 (13.8) 32 (19.9) 33 (13.5) 18 (20.2) 14 (19.4) 0.250
 Confusion (%) 47 (11.6) 17 (10.6) 33 (13.5) 9 (10.1) 8 (11.1) 0.519
 Neurological symptoms (%) 32 (7.9) 12 (7.5) 20 (8.2) 7 (7.9) 5 (6.9) 0.794
 Unconsciousness (%) 27 (6.7) 5 (3.1) 10 (4.1) 2 (2.2) 3 (4.2) 0.788
 Seizures (%) 13 (3.2) 3 (1.9) 9 (3.7) 1 (1.1) 2 (2.8) 0.315
Median latency, days (range)b 41 (1–10,000) 61 (2–10,000) 30.5 (1–3,650) 62 (2–743) 59 (3–10,000) 0.020
c
Recovery, days (range) 8 (2–112) 14 (3–112) 11 (2–97) 14 (3–98) 14 (3–112) 0.030
Hypertonic saline usage (%) 69 (17) 26 (16.1) 43 (17.6) 14 (15.7) 12 (16.7) 0.989
Tolvaptan (%)d 6 (1.5) 2 (1.2) 4 (2) 2 (2.2) 0 1.000.
Pontine myelinolysis (%) 5 (1.2) 5 (3.1) 0 2 (2.2) 3 (4.2) 0.081
Dead at discharge (%) 33 (8.1) 5 (3.1) 28 (11.5) 2 (2.2) 3 (4.2) 0.026
Number of drugs (%)
1 (%) 161 (39.8) 63 (39.1) 98 (40.2) 35 (39.3) 28 (38.9) 0.889    
2 (%) 151 (37.3) 63 (39.1) 88 (36.1) 36 (40.4) 29 (40.3) 0.665
3 (%) 62 (15.3) 21 (13.0) 41 (16.8) 11 (12.4) 10 (13.9) 0.551
4 (%) 22 (5.4) 10 (6.2) 12 (4.9) 5 (5.6) 5 (6.9) 0.524
5 (%) 9 (2.2) 4 (2.5) 5 (2.0) 1 (1.1) 3 (4.2) 1.0
Type of hyponatremia
SIADH (%) 121 (29.9) 56 (34.8) 65 (26.6) 31 (34.8) 25 (34.7) 0.220 ref.  
Average osmolality—mOsm/kg (SD) (sample = 121, SD) 246 (5.1) 245 (5.2) 247 (4.6) 246 (5.6) 243 (5.8) 0.741 ref.  
Average urine sodium concentration—mM 78.5 (29.1) 81.6 (33.9) 77.3 (25.7) 82.0 (35.0) 85.1 (34.8) 0.697 ref.  
(sample = 121, SD)
Average urine osmolality—mOsm/kg (sample = 121, SD) 413.7 (117.6) 445 (173.9) 404 (112.1) 444 (175.3) 446 (182.6) 0.591 ref  
Average urea concentration—mg/dL (sample = 121, SD) 17.1 (10.4) 15.1 (10.3) 17.9 (9.9) 15.9 (11.2) 14.3 (11.4) 0.460 ref.  
Serum creatinine—mg/dL (sample = 121, SD) 0.87 (0.38) 1.02 (0.34) 0.99 (0.21) 1.0 (0.43) 1.03 (0.40) 0.479 ref.  
2
eGFR (CKD-EPI) —mL/minute/1.73 m (sample = 121, SD) 93.7 (21.0) 90.0 (36.3) 95.6 (21.8) 89.0 (38.3) 92.1 (39.5) 0.251 ref  
Serum sodium concentration—mg/dL (sample = 121, SD) 112.3 (6.3) 111.9 (3.8) 117.9 (2.4) 111.2 (3.9) 112.7 (4.0) <0.001 ref.  
Serum chlorine concentration—mg/dL (sample = 121, SD) 101.4 (4.0) 103.5 (4.8) 104.7 (4.2) 104.7 (4.5) 100.6 (4.9) 0.763 ref.  

VOLUME 106 NUMBER 6 | DECEMBER 2019 | www.cpt-journal.com

(Continues)
 Table 3  (Continued)
Second period First period
Total very
Total ADR severe Total severe Very severe Very severe

Na (s) Na (s)
  < 122 mM Na < 116 mM 116–122 mM Na < 116 mM Na < 116 mM P value    
Hypovolemic hyponatremia (%) 284 (70.1) 105 (65.2) 179 (73.4) 58 (65.2) 47 (65.3) 0.222 <0.001  
 TIHe (%) 178 (43.9) 73 (45.3) 105 (43.1) 40 (44.9) 33 (45.8) 0.788 0.039 ref.
Average osmolality—mOsm/kg (sample = 117, SD) 253 (4.1) 251 (4.9) 256 (3.7) 250 (5.2) 253 (5.0) 0.861 0.083 ref.
Average urine sodium concentration—mM (sample = 117, 69.3 (16.3) 74.9 (19.1) 66.1 (14.0) 75.3 (19.7) 74.2 (19.9) 0.259 0.142 ref.
SD)
Average urine osmolality—mOsm/kg (sample = 178, SD) 416.9 (110.4) 431.6 (127.5) 409.2 (118.7) 431.8 (129.3) 431.3 (130.1) 0.329 0.245 ref.
Average urea concentration, mg/dL (sample = 178, SD) 32 (8.1) 29 (9.7) 33 (6.5) 30 (9.9) 27 (10.1) 0.875 <0.001 ref.
Serum creatinine—mg/dL (sample = 178, SD) 1.08 (0.23) 1.10 (0.34) 1.00 (0.24) 1.09 (0.39) 1.12 (0.37) 0.611 <0.001 ref.
eGFR (CKD-EPI)—mL/minute (sample = 178, SD) 62.5 (26.1) 64 (28.5) 66.1 (23.0) 63 (29.9) 66 (30.1) 0.791 <0.031 ref.
Serum sodium concentration—mg/dL (sample = 178, SD) 115.2 (5.3) 110.2 (3.3) 120.4 (1.8) 110.9 (3.5) 108.9 (3.7) <0.001 0.100 ref.
Serum chlorine concentration—mg/dL (sample = 178, SD) 80.2 (4.8) 73.0 (5.0) 85.6 (4.3) 72.5 (5.6) 73.7 (5.7) 0.045 <0.001 ref.

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 106 NUMBER 6 | DECEMBER 2019

 Non-TIH (%) 106 (26.2) 32 (19.9) 74 (30.3) 18 (20.2) 14 (19.4) 0.071 0.011 0.630
Average osmolality, mOsm/kg (SD) (sample = 77, SD) 266 (5.6) 264 (6.4) 265 (5.9) 265 (6.7) 263 (6.9) 0.903 <0.001 0.044
Average urine sodium concentration, mM (sample = 77, 54 (12.4) 52 (14.8) 51 (13.2) 53 (15.2) 51 (15.4) 0.644 0.014 0.024
SD)
Average urine osmolality, mOsm/kg (sample = 77, SD) 104 (22.2) 108 (29.5) 101 (19.5) 107 (29.9) 109 (30.0) 0.782 <0.001 <0.001
Average urea concentration, mg/dL (sample = 106, SD) 48 (11.1) 50 (12.7) 45 (9.3) 49 (13.4) 51 (13.5) 0.150 <0.001 0.046
Serum creatinine—mg/dL (sample = 106, SD) 1.08 (0.4) 1.10 (0.6) 1.05 (0.3) 1.10 (0.8) 1.11 (0.9) 0.591 0.032 0.247
eGFR (CKD-EPI) —mL/minute/1.73 m2 (sample = 106, SD) 60.3 (23.1) 59.2 (27.3) 62.1 (20.6) 59.0 (29.0) 59.4 (28.9) 0.791 0.010 0.650
Serum sodium concentration—mg/dL (sample = 106, SD) 114.9 (4.7) 110.8 (5.9) 121.3 (1.3) 110.3 (6.6) 111.2 (6.9) <0.001 0.432 0.469
Serum chlorine concentration—mg/dL (sample = 106, SD) 86.0 (5.2) 71.9 (5.0) 83.8 (4.7) 70.7 (5.8) 72.0 (5.4) 0.038 <0.001 0.619
Bold P values are statistically significant (P < 0.05).
CKD-EPI, chronic disease epidemiology collaborative; eGFR, estimated glomerular filtration rate; GFR, glomerular filtration rate; Na(s), serum sodium; ref., reference; SIADH, syndrome of inappropriate secretion of
antidiuretic hormone; TIH, thiazide-induced hyponatremia.
a
The six cases of children were during hospitalization. bThe period between the drug intake and the onset of symptoms. cThe period between the onset of symptoms or hyponatremia signal and serum sodium
back to normal range (≥ 135 mM). dAvailable from 2010. eTIH, by thiazide and thiazide-like diuretics, such as indapamide and chlortalidone.

1373
ARTICLE
ARTICLE

Figure 3  Predictive model based on a multivariable linear regression analysis of ADR-Na(s); the dependent variable was the inverse of serum
sodium (3 − (log-transformed) serum sodium), bootstrap estimation (10,000 iterations). ADR, adverse drug reaction; BCa, bias-corrected
and accelerated; F, the ratio of the mean regression sum of squares divided by the mean error sum of squares; R, correlation coefficient; R2,
determination coefficient; sig, signification; t, statistic obtained by dividing the regression coefficients between their corresponding typical
errors. Equivalent to the F statistic of the analysis of variance (ANOVA), t 2 = F. [Colour figure can be viewed at wileyonlinelibrary.com]

DISCUSSION intake or replacement, diuretic therapy, diarrhea and vomiting,

Incidence of hyponatremia and mineralocorticoid deficiency. In accordance with these re-
Hyponatremia is the most common electrolyte disturbance en- sults, we found that the causes of hypervolemic hyponatremia
countered in clinical practice and is associated with increased were not drug related. The causes of euvolemic hyponatremia
mortality, longer hospital stays, and increased institutionaliza- were predominantly neoplasia, central nervous system masses or
tion. 23 An accurate appreciation of the causes of hyponatremia hemorrhage, and medications. The main cause of hypovolemic
is essential for appropriate clinical management and also to pre- hyponatremia was medications.
vent the development of hyponatremia. At present, most pub- We found a significant association between drug-induced se-
lished reports had a retrospective design that relied on diagnoses vere hyponatremia and high outdoor temperatures. Jönsson et al.26
made by nonexpert clinicians retrospectively reviewing case found a relationship between outdoor temperatures during the
notes, frequently lacking sufficient detail to ensure an accurate summer (May–September) in Sweden and the reporting of sus-
diagnosis (i.e., documentation of volemic status and appropri- pected drug-induced hyponatremia and random controls to the
ate studies). 24 Our results have been obtained from a prospec- Medical Products Agency. Water intake and loss are approximately
tive program; alternative causes were confirmed. Soiza et  al. 25 40% higher in the summer than in the winter.27 Also, it is reported
found the main causes of hypervolemic hyponatremia were that salt appetite is not increased in summer heat.28 Medication
cardiac failure and cirrhosis and kidney diseases. The causes might affect sodium hemostasis and water hemostasis directly (e.g.,
of euvolemic hyponatremia were SIADH, medications, hypo- diuretics) and also antidiuretic hormone secretion centrally (e.g.,
thyroidism, primary polydipsia, and glucocorticoid deficiency; psychotropics), thereby increasing the risk of hyponatremia during
the causes of hypovolemic hyponatremia were inadequate fluid warm weather.

1374 VOLUME 106 NUMBER 6 | DECEMBER 2019 | www.cpt-journal.com

 ARTICLE

Table 4  Preventability of ADR-Na(s), STOPP criteria by periods: first period (from July 2007 to June 2012), second period
(from July 2012 to June 2017)
First Second
Total period period

STOPP criteria N (%) N (%) N (%)

Section A
A2 Any drug prescribed beyond the recommended duration, where treatment duration is well defined. 2 (3.0) 2 (3.0) 0 (0.0)
A3 Any duplicate drug class prescription, e.g., two concurrent NSAIDs, SSRIs, loop diuretics, ACE 10 (15.2) 3 (4.6) 7 (10.6)
inhibitors, anticoagulants.
Section B
B2 Verapamil or diltiazem with NYHA class III or IV heart failure (might worsen heart failure). 1 (1.5) 0 (0.0) 1 (1.5)
B5 Amiodarone as first-line antiarrhythmic therapy in supraventricular tachyarrhythmias (higher risk of 1 (1.5) 1 (1.5) 0 (0.0)
adverse effects than beta-blockers, digoxin, verapamil or diltiazem).
B7 Loop diuretic for dependent ankle edema without clinical, biochemical or radiological evidence of 18 (27.3) 5 (7.6) 13 (19.7)
heart failure, liver failure, nephrotic syndrome or renal failure (leg elevation and /or compression hosiery
usually more appropriate).
B8 Thiazide diuretic with current significant hypokalemia (i.e., serum K+ < 3.0 mmol/L), hyponatremia 8 (12.1) 2 (3.0) 6 (9.0)
(i.e. serum Na+ < 130 mmol/L), hypercalcemia (i.e. corrected serum calcium >2.65 mmol/L) or with
a history of gout (hypokalemia, hyponatremia, hypercalcemia and gout can be precipitated by thiazide
diuretic).
Section D
D2 Initiation of TCAs as first-line antidepressant treatment (higher risk of adverse drug reactions with 3 (4.6) 0 (0.0) 3 (4.6)
TCAs than with SSRIs or SNRIs).
D5 Benzodiazepines for ≥4 weeks (no indication for longer treatment; risk of prolonged sedation, confu- 2 (3.0) 2 (3.0) 0 (0.0)
sion, impaired balance, falls, road traffic accidents; all benzodiazepines should be withdrawn gradually
if taken for more than 4 weeks, given there is a risk of causing a benzodiazepine withdrawal syndrome if
stopped abruptly).
D9 Neuroleptic antipsychotic in patients with behavioral and psychological symptoms of dementia 3 (4.6) 0 (0.0) 3 (4.6)
(BPSD) unless symptoms are severe and other nonpharmacological treatments have failed (increased
risk of stroke).
D10 Neuroleptics, such as hypnotics, unless sleep disorder is due to psychosis or dementia (risk of con- 1 (1.5) 0 (0.0) 1 (1.5)
fusion, hypotension, extra-pyramidal side effects, falls).
D12 Phenothiazines as first-line treatment, given safer and more efficacious alternatives exist (pheno- 1 (1.5) 1 (1.5) 0 (0.0)
thiazines are sedative and have significant antimuscarinic toxicity in older people, with the exception
of prochlorperazine for nausea/vomiting/vertigo, chlorpromazine for relief of persistent hiccoughs and
levomepromazine as an antiemetic in palliative care).
Section H
H7 COX-2 selective NSAIDs with concurrent cardiovascular disease (increased risk of myocardial infarc- 1 (1.5) 0 (0.0) 1 (1.5)
tion and stroke).
Section J
J1 Sulphonylureas with a long duration of action (e.g. glibenclamide, chlorpropamide, glimepiride) with 1 (1.5) 1 (1.5) 0 (0.0)
type 2 diabetes mellitus (risk of prolonged hypoglycemia).
Section K
K1 Benzodiazepines (sedative, can cause reduced sensorium, impair balance). 3 (4.6) 2 (3.0) 1 (1.5)
K2 Neuroleptic drugs (can cause gait dyspraxia, Parkinsonism). 10 (15.2) 3 (4.6) 7 (10.6)
K4 Hypnotic Z-drugs e.g. zopiclone, zolpidem, zaleplon (can cause protracted daytime sedation, ataxia). 1 (1.5) 1 (1.5) 0 (0.0)
Total 66 (100) 23 (34.8) 43 (65.2)
ACE, angiotensin converting-enzyme; ADR-SS, drug-induced severe hyponatremia; NSAID, nonsteroidal anti-inflammataory drugs; NYHA, New York Heart
Association; SSRI, selective serotonin reuptake inhibitor; SNRI, serotonin and norepinephrine reuptake inhibitor; STOPP, Screening Tool of Older People’s
Prescriptions; TCA, tricyclic antidepressants.

Drugs implicated in drug-induced severe hyponatremia antihypertensives, and loop diuretics (72/786, 9.2%) causing
The leading causes of drug-induced severe or very severe hy- hypovolemic hyponatremia. Diuretics comprise one of the most
ponatremia were thiazides and thiazide-like drugs, such as in- common causes of hyponatremia in the elderly.29 Thiazides and
dapamide and chlortalidone (TIH, 178/786, 22.6% of cases), in thiazide-like diuretics have continued to be used as a first-line
fixed-dose combination with potassium-sparing diuretics or other treatment of hypertension in most European countries in 2018.30

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 106 NUMBER 6 | DECEMBER 2019 1375
ARTICLE

A systematic review and meta-analysis of all published cases of are typical of acute hyponatremia are triggered by cerebral edema
TIH showed that the mean patient age was 75 (95% CI, 73–77) and are associated with higher mortality (8.1%).41 In chronic hy-
years, 78% were women (95% CI, 74–82), and the latency to pre- ponatremia, the remaining cases, the degree of cerebral edema
sentation was 19.03 (95% CI, 7.97, 30.09) days, suggesting that was mild, thereby resulting in mild cognitive impairment but an
the recommended practice of performing a single measurement increased risk of falls, deterioration, and fatigue.7
of serum biochemistry 7–14  days after thiazide initiation might
be insufficient.31 These data are more extreme in this study, given Types of hyponatremia
TIH cases had a median (range) age of 81 (18–100) years, with The phenotype of drug-induced hypovolemic hyponatremia was
82.1% being women. similar to the criteria for SIADH outlined by Bartter and Schwartz,
The second-most frequent therapeutic subgroup was antide- including (i) decreased plasma osmolarity (<275  mOsm/kg),
pressants (73/786, 9.3%), followed by antiepileptics and anti- (ii) inappropriate urinary concentration (Uosm  >  100  mOsm/
psychotics in 68 cases (8.7%) causing SIADH. A single-center kg), (iii) euvolemia on clinical examination, (iv) elevated urinary
retrospective study of the distribution of etiologies for SIADH sodium (>20 mM), and (v) normal thyroid and adrenal function.42
showed an average age of drug-induced SIADH of 67.9 (SD, 16.9) These patients were also hypochloridemic, which is perhaps un-
years; 56.2% were women.32 The data from this study showed a surprising since the thiazide-sensitive chloride cotransporter
median (range) age of 76.5 (18–100) years; 65% were women. transports both sodium and chloride.43

Patient characteristics Predictive factors

Older patients were more prone to ADR-Na(s) because of im- To our knowledge, this is the first study to construct a model of
paired water-excretory capacity, which can mainly be attributed drug-induced severe hyponatremia. The results show that the
to age-related reduction in glomerular filtration rate  (GFR).33 model accurately predicts the level of drug-induced hyponatre-
Additionally, the decreased intrarenal generation of prostaglan- mia (the adjusted R 2 with bootstrap: 0.991). The proposed model
dins in advanced age might also impair the ability of elderly in- might be a useful tool in the evaluation of causality of severe hy-
dividuals to excrete water.34 Additionally, a higher sensitivity to ponatremia. The predictors associated with hyponatremia have
osmotic stimuli might occur in the geriatric population given already been assessed in other populations.44
that the relatively rare SIADH is more frequently observed in
elderly individuals.35 Estrogen and progesterone impact osmotic Preventability
regulation of arginine vasopressin and mediate sex differences in To our knowledge, this is also the first study to evaluate the use
blood pressure response to sodium loads. Because of this, young of potentially inappropriate medications among patients with
women are at greater risk of negative outcomes of hyponatremia. drug-induced hyponatremia. Inappropriate prescribing has been
As women age, lower estrogen and progesterone exposure also found to be associated with an increase in ADRs, poorer health-re-
increase arginine vasopressin, producing water retention.36 Our lated quality of life, hospital accidents, and emergency room visits
results are in concordance with the literature, and they are of in older patients.45 The study showed that 75.3% of drugs sus-
particular concern because the tolerance patients have to drugs pected of causing hyponatremia were prescribed to i­ndividuals
that can induce hyponatremia is reduced with aging.37 In addi- older than 65 years; 60.2% of cases received more than one of these
tion, an acute decrease in Na(s) can be superimposed on chronic drugs, and of them, 21.6% met some STOPP criteria. It is possible
hyponatremia (latency up to 10,000 days). The 22.9% of the pa- to identify a subgroup of patients at increased risk of severe hypo-
tients in the severe hyponatremia group and 50% of the patients natremia based on age, gender, outdoor temperatures, and medica-
with very severe hyponatremia had dementia, inadequate fluid tion that induced hyponatremia and for whom to select alternative
intake, and hyponatremic medication, which is highly prev- treatment or to focus additional monitoring.
alent in this population, 38 so this combination might explain
the severity of hyponatremia in patients with dementia. On the Strengths and limitations of PPLSH
other hand, hyponatremia has been recently associated with the The prospective design of the PPLSH permits the early identi-
occurrence of dementia. In a population-based retrospective fication of potential ADRs and discussion of these ADRs with
study, hyponatremia patients had a 2.36-fold higher chance of clinicians, thereby improving the diagnosis and subsequent man-
suffering dementia, with a dose–response relationship.39 agement. The methodological limits of the study are our data
sources. It is possible that some ADR-Na(s) cases were missed
Manifestations of drug-induced hyponatremia during the process of attributing alternative causes through the
In our results, falls were the most frequent symptom of severe EMR review (phase II). Although such a possibility is low, given
and very severe hyponatremia; of these, 58.6% (82/140) were se- that the alternative causes presented are well known, we cannot
rious falls. We found fewer fractures or brain injuries in the very exclude the possible underestimation of ADR-Na(s).
severe hyponatremia group compared with the severe hyponatre-
mia group (27.5% vs. 9.3%, P < 0.001), probably due to a strong In conclusion
association between immobility, dementia, and very severe hy- Severe hyponatremia is often due to medication. Clinical aware-
ponatremia.40 Neurological symptoms, unconsciousness, and ness should be increased, especially in older female patients using
seizures were observed in 29.4% of the patients. These symptoms cardiovascular or nervous system drugs.

1376 VOLUME 106 NUMBER 6 | DECEMBER 2019 | www.cpt-journal.com

 ARTICLE

METHODS was categorized as very severe hyponatremia, and Na(s) 116–122  mM

Setting was categorized as severe hyponatremia. The Anatomical Therapeutic
La Paz University Hospital in Madrid, Spain, is a tertiary care teaching Chemical Classification of drugs, groups, and subgroups was used to
facility. From July 2007 to June 2017, all Na(s) laboratory results at ad- classify the related drugs of ADR-Na(s). Indications for medication were
mission or during hospitalization were monitored by the PPLSH. Data evaluated according to STOPP/Screening Tool to Alert doctors to the
from patients who died in the emergency ward were also included. The Right Treatment criteria. 20
program was conducted in accordance with the Spanish Personal Data
Protection Law.46 Approval to publish the program was obtained from Variables included in the analysis
the Institutional Review Board at La Paz University Hospital, protocol Meteorological data were obtained from the open data repository at
PI-3226. All ADRs were reported to the Spanish pharmacovigilance the Spanish State Meteorological Agency (AEMET: https​ ://datos​
system. clima.es/Aemet​histo​rico/Meteo​stati​on.php). The following variables
were recorded in the population with hyponatremia: age, sex, date, and
Signal-sodium definition maximum and minimum outdoor temperature of the sodium signal
Cases of Na(s) <  116  mM were recorded during the first period (July day. The following variables were also recorded in the population with
2007–June 2012). Cases of Na(s) < 122 mM were recorded during the sec- ADR-Na(s): community acquired or hospital acquired, underlying con-
ond period (July 2012–June 2017). Therefore, cases of Na(s) < 116 mM ditions, main symptom, latency, recovery, hypertonic saline usage, pon-
were recorded over the decade, and cases of Na(s) 116–122 mM were re- tine myelinolysis, dead at discharge, number of related drugs per patient,
corded over the last 5 years. These values correspond to grade 4 and grade indication of the medication, groups and subgroups of the Anatomical
3, respectively, of the World Health Organization Toxicity Grading Scale Therapeutic Chemical Classification, type of hyponatremia (SIADH or
for determining the severity of adverse drug events.47 hypovolemic hyponatremia), serum osmolality, urine sodium concen-
tration, urea concentration, GFR estimated by chronic disease epidemi-
Information system used and coverage ology collaborative method, Na(s) concentration, and serum chlorine
A specific database application was developed within the Integrated concentration.
Laboratory System (LABTrack Integrated Laboratory System, develop-
ment version; TrackHealth, Woolloomooloo, Australia) to detect sys- Sample size calculation
tematically predefined automatic laboratory signals. At the time of the Accepting an alpha risk of 0.05 and a beta risk of 0.2 in a two-sided test,
study, EMRs included all laboratory data, images and other exploratory 332 participants in the ADR-Na(s) group and 996 in the non–ADR-
results, previous medical reports, and discharge summaries. Na(s) group were necessary to recognize as statistically significant an
odds ratio (OR) greater than or equal to 2. A minimum of 126 partici-
ADR definition pants were necessary in each group of ADR-Na(s) to find as statistically
The ICH, International Conference on Harmonisation, E2D definition significant a proportion difference of 0.12 (arcsin approximation).
of ADR was used.48 For the program’s purpose, ADRs caused by acciden-
tal or intentional overdoses were excluded. Medical errors, considered to Statistical analysis
be any error in written prescription, dispensation, or administration were The in-hospital incidence rates of ADR-Na(s) and of other etiologies
excluded; errors in decision making (used in contraindicated clinical con- were calculated by dividing the number of automatic laboratory signals
ditions or drug interactions) were considered ADRs and therefore were by the number of admissions during the study period. Uncertainty of es-
included. timation was assessed by calculation of the two-sided Poisson 95% con-
fidence interval. The χ2 test was performed to compare sex distribution
Procedure for ADR detection, evaluation, and classification and discrete variables, and Student’s t-test or the Mann–Whitney U-test,
The procedure of PPLSH has been described elsewhere. 22 Briefly, in as appropriate, was used to compare continuous variables. We built a re-
phase I, on-file laboratory data at admission or during the hospital ceiver operating characteristic curve to evaluate the seasonality, age, sex,
stay were screened 7 days a week, 24 hours per day, for Na(s) signal. In and serum sodium level of the patients with hyponatremia by drugs vs.
phase II, the patients were identified to avoid duplicates, and EMRs hyponatremia by other causes using R Core Team (2018) (R Foundation
were reviewed. In phase III, the remaining cases, case-by-case studies for Statistical Computing, Vienna, Austria). The k-fold cross-validation
were performed. The cause and the type of hyponatremia were estab- of the logistic regression model was done in k sets (A, B, and C). One by
lished according to the judgment of expert clinicians (Supplementary one, a set is selected as test set. Then, one by one, one of the remaining
Information). The causality assessment of drugs was performed using sets is used as a validation set; 1,000 simulations have been used. A pre-
an algorithm of causality (Table S1b). dictive model based on a multivariable linear regression analysis of ADR-
Na(s) was constructed, being the dependent variable of serum sodium
concentration on the signal day in the ADR group (3 − (log-transformed
Collection of patient data and reporting Na(s))). A bootstrap estimation of the standard error and CIs for R 2 was
For all patients initially categorized as having an ADR, a complete ad- performed based on 10,000 samples using IBM SPSS Statistics version
verse reaction report was submitted to the pharmacovigilance center in 20.0 (IBM Corporation, Armonk, NY). The models were constructed
Madrid. with adult patients (first hyponatremia), and valued covariates were
log-transformed.
Diagnostic definition
Pseudohyponatremia was defined as a decreased Na(s) concentration
SUPPORTING INFORMATION
that does not correspond to a real hypotonic disorder due to hyperlip- Supplementary information accompanies this paper on the Clinical
idemia, hyperproteinemia, or hyperglycemia. Preanalytical errors were Pharmacology & Therapeutics website (www.cpt-journal.com).
defined as blood samples extracted and diluted with dextrose. Diagnoses
were coded according to the International Classification of Diseases, Supplementary Information. Procedure for ADR detection and
Ninth Revision, Clinical Modification codes.49 A GFR < 60 mL/min- evaluation.
ute/1.73  m2  for 3  months is classified as chronic kidney disease, irre- Table S1. (a) Evaluation of volume status in hyponatremia. (b) Algorithm
spective of the presence or absence of kidney damage.50 Na(s) < 116 mM of the Spanish Pharmacovigilance System.

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 106 NUMBER 6 | DECEMBER 2019 1377
ARTICLE

ACKNOWLEDGMENTS hyponatraemia. Eur. J. Endocrinol. 170, G1–47 (2014). Erratum

The authors would like to thank Juliette Siegfried and her team at in: Eur. J. Endocrinol.171(1):X (2014). https​://doi.org/10.1530/
ServingMed.com for their editing of the manuscript. eje-13-1020
16. Greenberg, A., et al. Current treatment practice and outcomes.
Report of the hyponatremia registry. Kidney Int. 88, 167–177 (2015).
FUNDING
17. Peri, A. Management of hyponatremia: causes, clinical aspects,
The study was performed as part of the routine work of the Functional
differential diagnosis and treatment. Expert Rev. Endocrinol.
Unit of Risk Management of La Paz University Hospital.
Metab. 14, 13–21 (2019).
18. Spinewine, A. et al. Appropriate prescribing in elderly people: how well
CONFLICT OF INTEREST can it be measured and optimised? Lancet 370, 173–184 (2007).
The authors declared no competing interests for this work. 19. Gallagher, P. & O'Mahony, D. STOPP (Screening Tool of Older
Persons’ potentially inappropriate Prescriptions): application to
acutely ill elderly patients and comparison with Beers’ criteria.
AUTHOR CONTRIBUTIONS Age Ageing 37, 673–679 (2008).
E.R., A.R., and J.F. wrote the manuscript. E.R. and J.F. designed the 20. O'Mahony, D., O'Sullivan, D., Byrne, S., O'Connor, M.N., Ryan, C.
research. A.R., J.Q., I.G., L.D., L.M., R.M., H.Y.T., and M.M. performed & Gallagher, P. STOPP/START criteria for potentially inappropriate
the research. E.R., A.R., and J.Q. analyzed the data. J.C.M., A.M.B., and prescribing in older people: version 2. Age Ageing 44, 213–218
A.J.C. contributed new reagents/analytical tools. (2015).
21. Hannan, T.J. Detecting adverse drug reactions to improve patient
© 2019 The Authors Clinical Pharmacology & Therapeutics © 2019 outcomes. Int. J. Med. Inform. 55, 61–64 (1999).
American Society for Clinical Pharmacology and Therapeutics 22. Ramírez, E., et al. A pharmacovigilance program from laboratory
signals for the detection and reporting of serious adverse drug re-
actions in hospitalized patients. Clin. Pharmacol. Ther. 87, 74–86
1. Liamis, G., Rodenburg, E.M., Hofman, A., Zietse, R., Stricker, (2010).
B.H. & Hoorn, E.J. Electrolyte disorders in community subjects: 23. Woodward, M., Gonski, P., Grossmann, M., Obeid, J., Scholes, R.
prevalence and risk factors. Am. J. Med. 126, 256–263 & Topliss, D.J. Diagnosis and management of hyponatraemia in
(2013). the older patient. Intern. Med. J. 48 (suppl. 1), 5–12 (2018).
2. Giordano, M. et al. Seasonal variations of hyponatremia in the 24. Hannon, M.J. & Thompson, C.J. The syndrome of inappropriate
emergency department: age-related changes. Am. J. Emerg. Med. antidiuretic hormone: prevalence, causes and consequences. Eur.
35, 749–752 (2017). J. Endocrinol. 162 (suppl. 1) , S5–S12 (2010).
3. Hoyle, G.E., Chua, M. & Soiza, R.L. Prevalence of hyponatremia 25. Soiza, R.L., Hoyle, G.E. & Chua, M.P.W. Electrolyte and salt distur-
in elderly patients. J. Am. Geriatr. Soc. 54, 1473, author reply bances in older people: causes, management and implications.
1473–1474 (2006). Rev. Clin. Gerontol. 18, 143–158 (2008).
4. Miller, M., Morley, J.E. & Rubenstein, L.Z. Hyponatremia in a nurs- 26. Jönsson, A.K., Lövborg, H., Lohr, W., Ekman, B. & Rocklöv, J.
ing home population. J. Am. Geriatr. Soc. 43, 1410–1413 (1995). Increased risk of drug-induced hyponatremia during high tempera-
5. Edelman, I.S., Leibman, J., O'Meara, M.P. & Birkenfeld, L.W. tures. Int. J. Environ. Res. Public Health. 14, 827 (2017). https​://
Interrelations between serum sodium concentration, serum doi.org/10.3390/ijerp​h1407​0827.
osmolarity and total exchangeable sodium, total exchangeable 27. Malisova, O., Bountziouka, V., Panagiotakos, D.Β., Zampelas, A. &
potassium and total body water. J. Clin Invest. 37, 1236–1256 Kapsokefalou, M. Evaluation of seasonality on total water intake,
(1958). water loss and water balance in the general population in Greece.
6. Adrogué, H.J. & Madias, N.E. The challenge of hyponatremia. J. J. Hum. Nutr. Diet. 26 (suppl. 1), 90–96 (2013).
Am. Soc. Nephrol. 23, 1140–1148 (2012). 28. Leshem, M. Salt appetite is not increased in summer heat.
7. Renneboog, B., Musch, W., Vandemergel, X., Manto, M.U. & Appetite 108, 28–31 (2017).
Decaux, G. Mild chronic hyponatremia is associated with falls, 29. Clayton, J.A., Rodgers, S., Blakey, J., Avery, A. & Hall, I.P. Thiazide
unsteadiness, and attention deficits. Am. J. Med. 119(71), e1–e8 diuretic prescription and electrolyte abnormalities in primary care.
(2006). Br. J. Clin. Pharmacol. 61, 87–95 (2006).
8. Hoorn, E.J. et al. Mild hyponatremia as a risk factor for frac- 30. Mancia, G. et al. 2013 ESH/ESC guidelines for the management
tures: the Rotterdam Study. J. Bone Miner. Res. 26, 1822–1828 of arterial hypertension: the Task Force for the management of
(2011). arterial hypertension of the European Society of Hypertension
9. Sajadieh, A., Binici, Z., Mouridsen, M.R., Nielsen, O.W., Hansen, (ESH) and of the European Society of Cardiology (ESC). Eur. Heart
J.F. & Haugaard, S.B. Mild hyponatremia carries a poor prognosis J. 34, 2159–2219 (2013).
in community subjects. Am. J. Med. 122, 679–686 (2009). 31. Barber, J., et al. A systematic review and meta-analysis of thia-
10. Gankam Kengne, F. & Decaux, G. Hyponatremia and the brain. zide-induced hyponatremia: time to reconsider electrolyte moni-
Kidney Int. Rep. 3, 24–35 (2017). toring regimens after thiazide initiation? Br. J. Clin. Pharmacol. 79,
11. Corona, G., Giuliani, C., Verbalis, J.G., Forti, G., Maggi, M. & 566–577 (2015).
Peri, A. Hyponatremia improvement is associated with a reduced 32. Shepshelovich, D., Schechter, A., Calvarysky, B., Diker-Cohen, T.,
risk of mortality: evidence from a meta-analysis. PLoS One 10, Rozen-Zvi, B. & Gafter-Gvili, A. Medication-induced SIADH: distri-
e0124105 (2015). bution and characterization according to medication class. Br. J.
12. Vaghasiya, R.P., DeVita, M.V., Panagopoulos, G. & Michelis, M.F. Clin. Pharmacol. 83, 1801–1807 (2017).
Hyponatremic patients demonstrate improved cognition after 33. Frenkel, N.J. et al. Thiazide-induced hyponatraemia is associated
an increase in serum sodium. NKF Annual Meeting, Orlando, FL, with increased water intake and impaired urea-mediated water ex-
USA, April 6–10, 2010. Poster 304. cretion at low plasma antidiuretic hormone and urine aquaporin-2.
13. Soiza, R.L. & Talbot, H.S.C. Management of hyponatraemia in J. Hypertens. 33, 627–633 (2015).
older people: old threats and new opportunities. Ther. Adv. Drug 34. Clark, B.A., Shannon, R.P., Rosa, R.M. & Epstein, F.H. Increased
Saf. 2, 9–17 (2011). susceptibility to thiazide-induced hyponatremia in the elderly. J.
14. Cumming, K., Hoyle, G.E., Hutchison, J.D. & Soiza, R.L. Am. Soc. Nephrol. 5, 1106–1111 (1994).
Bioelectrical impedance analysis is more accurate than clinical 35. Helderman, J.H., Vestal, R.E., Rowe, J.W., Tobin, J.D., Andres, R.
examination in determining the volaemic status of elderly patients & Robertson, G.L. The response of arginine vasopressin to intra-
with fragility fracture and hyponatraemia. J. Nutr. Health Aging. venous ethanol and hypertonic saline in man: the impact of aging.
18, 744–750 (2014). J. Gerontol. 33, 39–47 (1978).
15. Spasovski, G. et al.; Hyponatraemia Guideline Development 36. Stachenfeld, N.S. Hormonal changes during menopause and the
Group. Clinical practice guideline on diagnosis and treatment of impact on fluid regulation. Reprod Sci. 21, 555–561 (2014).

1378 VOLUME 106 NUMBER 6 | DECEMBER 2019 | www.cpt-journal.com

 ARTICLE

37. Filippatos, T.D., Makri, A., Elisaf, M.S. & Liamis, G. Hyponatremia dwelling older patients. Br. J. Clin. Pharmacol. 77, 201–210
in the elderly: challenges and solutions. Clin. Interv. Aging 14, (2014).
1957–1965 (2017). 46. Organic law 15/1999 of 13 December 1999 on the Protection
38. Timmons, S., et al. Dementia in older people admitted to hospi- of Personal Data (BOE núm. 298, de 14-12-1999, pp. 43088-
tal: a regional multi-hospital observational study of prevalence, 43099). <http://apdcat.gencat.cat/web/.conte​nt/01-autor​itat/
associations and case recognition. Age Ageing 44, 993–999 norma​tiva/docum​entos/​960.pdf> (1999). Accessed February 4,
(2015). 2019.
39. Chung, M.C. et al. Hyponatremia and increased risk of dementia: 47. World Health Organization. WHO Handbook for Reporting Results
a population-based retrospective cohort study. PLoS One 12, of Cancer Treatment. WHO Offset publication No. 48 (World
e0178977 (2017). Health Organization, Geneva, Switzerland, 1979). Section VIII:
40. Cooper, C., Sommerlad, A., Lyketsos, C.G. & Livingston, G. WHO Toxicity Grading Scale for determining the severity of
Modifiable predictors of dementia in mild cognitive impairment: adverse events: 132–135, February 6, 2003. <http://www.
a systematic review and meta-analysis. Am. J. Psychiatry 172, icssc.org/Docum​ents/Resou​rces/AEMan​ual20​03App​endic​esFeb​
323–334 (2015). ruary_06_2003f​inal.pdf> (2003). Accessed February 4,
41. Fraser, C.L. & Arieff, A.I. Epidemiology, pathophysiology, and 2019.
management of hyponatremic encephalopathy. Am. J. Med. 102, 4 8. ICH Guideline on E2D Postapproval Safety Data Management:
67–77 (1997). Definitions and Standards for Expedited Reporting. London, May
42. Bartter, F.C. & Schwartz, W.B. The syndrome of inappropriate 2004. <http://eudra​vigil​ance.ema.europa.eu/human/​ichAn​dEtra​
secretion of antidiuretic hormone. Am. J. Med. 42, 790–806 nsmis​sion03.asp> (2003). Accessed July 17, 2019.
(1967). 49. World Health Organization. International Classification of
4 3. Ware, J.S. et al. Phenotypic and pharmacogenetic evaluation of Diseases, Ninth Revision (ICD-9). (World Health Organization,
patients with thiazide-induced hyponatremia. J. Clin. Invest. 127, Geneva, Switzerland, 1977).
3367–3374 (2017). 50. National Kidney Foundation. K/DOQI clinical practice guidelines
4 4. Hawkins, R.C. Age and gender as risk factors for hyponatremia for chronic kidney disease: evaluation, classification and stratifi-
and hypernatremia. Clin. Chim. Acta 337, 169–172 (2003). cation. Am. J. Kidney Dis. 39(suppl. 1), S1–S266 (2002). <https​://
45. Cahir, C., Bennett, K., Teljeur, C. & Fahey, T. Potentially inappro- www.kidney.org/sites/​defau​lt/files/​docs/ckd_evalu​ation_class​
priate prescribing and adverse health outcomes in community ifica​tion_strat​ifica​tion.pdf> (2002). Accessed February 4, 2019.

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 106 NUMBER 6 | DECEMBER 2019 1379