Sage Journals: Discover world-class research

Abstract

Introduction:

Thiazide diuretics are recommended as first-line therapy for hypertension in older adults. However, thiazides are also associated with hyponatraemia-related hospitalisations in older patients. This study aims to determine the predictors of hospitalisation due to thiazides usage in older adults.

Methods:

This is a retrospective matched case-control study. Patients aged ⩾65 admitted due to adverse drug reactions based on International Classification of Diseases, Ninth Revision (ICD9) codes from the period of 1 June to 31 December 2011 in Singapore General Hospital were extracted. Patients with the ICD9 code E944.3 Saluretics causing adverse effects in therapeutic use and who experienced thiazide-induced hyponatraemia were identified. Controls were identified from a pool of patients from outpatient clinics who were prescribed thiazide during the study period. Each case was matched to four controls based on gender and race. Patients’ demographics, length of stay, and cost of hospital admission were obtained.

Results:

In total, 19 cases with thiazide-induced hyponatraemia were matched with 76 controls. Cases were older than control (78.8±6.1 vs. 75.6±7.0, p=0.052), with the majority being females (84.2%) and Chinese (94.7%). The mean length of stay was 4 (±3) days; the mean cost of stay was SGD 1118 (±898). Serum potassium levels and concurrent use of beta-blockers were identified as unadjusted possible predictors for hospitalisation due to thiazide-induced hyponatraemia.

Conclusion:

Potential predictors of hospitalisation due to thiazide-induced hyponatraemia include low potassium levels and concurrent use of beta-blockers. Identification of predictors is crucial to guide safe and effective prescribing of thiazides in older patients.

Keywords

Thiazide diuretics older adults hospitalisation adverse drug events

Introduction

According to the Eighth Joint National Committee (JNC) and The National Institute of Health and Care Excellence (NICE), thiazide diuretics are one of the four main classes of antihypertensive recommended as first-line therapy for hypertension, besides angiotensin-converting enzyme inhibitors, angiotensin receptor blockers (ARB) and calcium channel blockers.^1,2 Thiazides are routinely used agents for older adults and their use was recommended as the preferred initial treatment of hypertension in the seventh report of JNC.³ The antihypertensive efficacy, simplicity of regimen and low cost, as well as the morbidity and mortality benefits seen in patients with high risk of cardiovascular disease and stroke, contributes to their place in the treatment of hypertension.^3–5

Despite being an effective antihypertensive, compared with loop diuretics, thiazides are prone to cause metabolic side effects such as hyponatraemia in older adults.^6–11 Studies have reported cases of hyponatraemia occurring in 6.9–33% of older patients.^7–9 Depending on the predisposing risk factors, late-onset thiazide-induced hyponatraemia has been reported months and years after initiation of the therapy.¹⁰ Reduction of free water clearance by the kidneys, increased excretion of sodium, and stimulation of antidiuretic hormone secretion due to reduced intravascular volume are some of the mechanisms that contribute to thiazide-induced hyponatraemia.^10,11 Patient factors such as the reduced excretion of free water due to age-related decline in renal function, high fluid intake, and low sodium intake can also contribute to hyponatraemia.¹⁰ Severe hyponatraemia is associated with neurological deficits, such as cerebral oedema and central pontine myelinosis, which could increase the risk of falls, seizures, and coma.^10,12,13

In view of significant morbidity from thiazide-induced hyponatraemia in older patients, preventive measures should be taken to guide safe and effective prescribing of thiazides.^10,12,13 Therefore, identification of predictors for hospitalisation is crucial for older patients who are susceptible to thiazide-induced hyponatraemia. Older patients are also at increased risk of adverse drug events as a result of their advanced age, polypharmacy, and multiple comorbidities.^14,15 A retrospective, case-control study conducted in Hong Kong reported that serum potassium, use of indapamide, elderly home institutionalisation and physical immobility were risk factors for thiazide-induced hyponatraemia.¹⁶ There are currently no published studies that have evaluated the predictors of hospitalisations due to thiazide-induced hyponatraemia in older Singaporeans.

The objective of this study is to explore the predictors of hospitalisations due to thiazide-induced hyponatraemia in older adults.

Methods

This was a retrospective matched case-control study conducted in the Singapore General Hospital (SGH), Singapore. The study protocol was reviewed and approved by the SGH Centralized Institution Research Board. Older adults, aged 65 and above, admitted to SGH for thiazide-induced hyponatraemia within the period of 1 June 2011 to 31 December 2011 were included in the study. The patients were identified using the specific ICD 9 primary diagnosis code, ICD-9-CM E944.3 Saluretics causing adverse effects in therapeutic use, extracted electronically from clinical manager software (Citrix™) and Hospital Inpatient Discharge Summary (HIDS). Saluretics, which include thiazide diuretics, are agents that facilitate the renal excretion of sodium. For the purpose of the study, hyponatraemia was defined as symptomatic with a plasma sodium level of <130 mmol/l necessitating hospitalisation.^17,18

The controls selected were patients from the outpatient setting who were prescribed and dispensed with a thiazide diuretic during the study period of 1 June 2011 to 31 December 2011. The controls were identified from the dispensing software, Maxcare, used in SGH. Each case was matched to four controls to increase the statistical power of the small sample size.¹⁹ The cases were matched to their gender and race, neither of which being found to be associated with thiazide-induced hyponatraemia.^7,11,16 Cases were matched to the list of controls and four controls per case were selected using the random number function in Microsoft Excel 2011 Version 14.2.0.

Information such as baseline characteristics, biochemical characteristics, medical conditions and concurrent medication use were extracted from HIDS and Citrix™. The biochemical lab values extracted for cases were at the time of admission and for the controls were the first available lab parameters within the study period. Additional information on other Adverse Drug Reaction (ADR) related admissions such as the types of ADR, the length of stay, and the cost of admission were obtained.

Statistical analysis

Descriptive statistics were used to report demographic characteristics. Fisher’s exact test and t-test were used to compare categorical and continuous variables, respectively, between cases and controls. Variables with p-values <0.2 in univariate analyses were included in a logistic regression model to determine the magnitude and direction of their associations with thiazide-related admissions. The analyses were performed using SPSS Statistics for Windows, Version 17.0 (Chicago: SPSS Inc.) at the 5% significance level.

Results

During the study period, 134 older patients were admitted to SGH for ADR. Saluretics (16.4%) were the second most causative drugs contributing to ADR-related admission. Based on the ICD9 codes, 22 patients admitted due to adverse effects from saluretics were identified, but three patients were admitted due to thiazide-induced hypokalaemia and were therefore excluded from the study. Of the remaining 19 patients with thiazide-induced hyponatraemia who were included in the analysis, two (10.5%) experienced both hyponatraemia as well as hypokalaemia. The causative agent for all cases was hydrochlorothiazide. Upon admission, hydrochlorothiazide was withdrawn and hyponatraemia resolved subsequently. The mean lengths of stay (LOS) for admission due to saluretics is 4 (±3) days with a cost of SGD $1118.45 (±898.00).

Table 1 describes the characteristics of patients for both controls and cases. Of the 19 patients in the case group, 16 (84.2%) were female. Chinese (n=18, 94.7%) patients accounted for the majority of the case group. Cases were older than controls (78.8±6.1 vs. 75.6±7.0 years, p=0.052). The mean serum sodium values for the case and control were 121.6 (±4.7) and 137.6 (±3.9) mmol/l, respectively.

Table 1.

Baseline characteristics of the Study Participants.

Baseline characteristics	Case group	Control group	p-value	95% CI
Female, n (%)	16 (84.2)	64 (84.2)	0.619	0.252–3.970
Race, n (%)
Chinese	18 (94.7)	72 (94.7)
Malay	1 (5.3)	4 (5.3)	0.681	0.105–9.501
Age, mean (±SD) years	78.84 (±6.11)	75.58 (±6.96)	0.052	(-0.026) – 6.559
Serum creatinine, mean (±SD) mmol/l	95.58 (±55.39)	146.24 (±91.29)	0.004	(−83.703)–(−17.403)
Serum sodium, mean (±SD) mmol/l	121.58 (±4.74)	137.62 (±3.91)	0.000	(−18.339)–(−12.688)
Serum potassium, mean (±SD) mmol/l	3.99(±0.66)	4.36 (±0.58)	0.034	(−0.7108)–(−0.0313)
Total medication, mean (±SD)	7.84 (±3.72)	8.09 (±3.11)	0.789	(−2.154)–1.654
Total comorbidities, mean (±SD)	3.89 (±1.73)	4.70 (±1.68)	0.181	(−1.510)–0.299
Types of comorbidities, n (%)
Hypertension	18 (94.7)	71 (93.4)	0.655	0.139–11.537
Diabetes	9 (47.4)	45 (59.2)	0.250	0.226–1.702
IHD	4 (21.1)	16 (21.1)	0.636	0.291–3.432
Heart Failure	1 (5.3)	3 (3.9)	0.597	0.133–13.773
Atrial fibrillation	1 (5.3)	4 (5.3)	0.740	0.105–9.501
Previous MI	1 (5.1)	1 (1.3)	0.362	0.249–69.843
Other CVD	3 (15.8)	5 (6.6)	0.196	0.576–12.304
Hyperlipidaemia	13 (68.4)	43 (56.6)	0.251	0.571–4.840
Renal impairment	1 (5.3)	22 (28.9)	0.024	0.017–1.085
Concurrent medications, n (%)
ACE inhibitors	10 (52.6)	33 (43.4)	0.320	0.528–3.968
ARB inhibitors	7 (36.8)	32 (42.1)	0.442	0.284– 2.263
Calcium channel blocker	10 (52.6)	44 (57.9)	0.436	0.295– 2.217
Beta blocker	15 (78.9)	39 (51.3)	0.025	1.081– 11.708
Nitrates	3 (15.8)	15 (19.7)	0.491	0.196– 2.960
Alpha-adrenergic blocker	2 (10.5)	5 (6.6)	0.426	0.298– 9.358
Anti-anginal (trimetazidine)	1 (5.3)	3 (3.9)	0.597	0.133– 13.773
Statin	13 (68.4)	55 (72.4)	0.467	0.278–2.461
Loop diuretic	2 (10.5)	5 (6.6)	0.426	0.298–9.358
Biguanide	5 (26.3)	13 (17.1)	0.269	0.530–5.649
Sulphonlyurea	7 (36.8)	21 (27.6)	0.301	0.530–4.406
Dipeptidyl peptidase 4	1 (5.3)	2 (2.6)	0.492	0.176–23.942
Proton pump inhibitor	6 (31.6)	29 (38.2)	0.401	0.256–2.186
H2 receptor blocker	1 (5.3)	5 (6.6)	0.655	0.087–7.180
Aspirin	7 (36.8)	21 (27.6)	0.301	0.530–4.406
Clopidogrel	2 (10.5)	6 (7.9)	0.503	0.254–7.407
Ticlopidine	1 (5.3)	1 (1.3)	0.362	0.249–69.843
Theophylline	1 (5.3)	1 (1.3)	0.362	0.249–69.843
Antihistamine	1 (5.3)	2 (2.6)	0.492	0.176–23.942
Lactulose	2 (10.5)	3 (3.9)	0.260	0.443–18.490
Senna	3 (15.8)	2 (2.6)	0.053	1.070–44.968
Levothyroxine	1 (5.3)	1 (1.3)	0.362	0249–69.843

CI: confidence interval; SD: standard deviation; MI: myocardial infarction; CVD: cardiovascular disease.

Serum creatinine levels for cases were lower than controls, (95.6±55.4 vs. 146.2±91.3 mmol/l, p<0.05). The potassium levels were within normal range for both cases and control, with lower levels in the cases than in the controls (3.99±0.66 and 4.36±0.58 mmol/l, p<0.05). The serum creatinine and potassium levels were noted to be significantly associated with thiazide-induced hyponatraemia. Total number of concurrent medications and other comorbidities were not associated with thiazide-induced hyponatraemia. However, the concurrent use of beta-blockers was found to increase the risk of experiencing thiazide-induced hyponatraemia by 3.5-fold (odds ratio: 3.56, 95% confidence interval [CI] 1.08–11.71). Other concurrent medications used were not found to be associated with thiazide-induced hyponatraemia. None of the patients in this study were prescribed any drugs that may affect the central nervous system, which are deemed to be a common cause of hyponatraemia in older adults.²⁰ Multivariate logistic regression of variables with p<0.2 in univariate analyses did not identify any significant predictors for thiazide-induced hyponatraemia-related hospitalisations.

Discussion

Due to polypharmacy, age-related pharmacodynamics and pharmacokinetic changes, older patients tend to experience more incidence of ADRs.²¹ In our study, 134 ADR-related admissions of older patients in a period of 6 months were reported. Hydrochlorothiazide was the second most common causative agent. Thiazide-induced hyponatraemia posed a significant burden to the healthcare system with incurred LOS of 4 (±3) days, and cost of hospitalisation of SGD $1118.45 (±898.00). These results are consistent with other studies, which showed that hyponatraemia-related hospitalisation was a significant predictor of medical costs.²² It was also associated with increased LOS and cost of stay in hospitalised patients.²³ Callahan et al. reported that in 2047 patients with hyponatraemia the mean LOS was 8 days, at cost USD $16,606.²³ The reduced cost of hospital care noted in our study is probably due to the shorter LOS and lower cost of care in Singapore.

This study found that low serum potassium level (K⁺<3.5mmol/l) was a possible predictor of hospitalisation due to thiazide-induced hyponatraemia, which corroborates with another study by Chow et al.¹⁶ The occurrence of low potassium and thiazide-induced hyponatraemia was due to the combination of kaliuretic effects and the hyponatraemic effects of thiazide through the sodium–potassium transcellular ion exchange. While evaluating hospitalisations for thiazide-induced hyponatraemia, Rastogi et al. also found that hypokalaemia was one of the most associated significant factors.²⁴

The serum creatinine in our study population was found to be associated with thiazide-induced hyponatraemia, suggesting that patients with high serum creatinine are less likely to experience thiazide-induced hyponatraemia. However, serum creatinine was not found to be a significant predictor of hospitalisation due to thiazide-induced hyponatraemia in other studies.^16,25,26 Contrary to our results, age-related decline in renal function is a possible contributory factor for hyponatraemia and may be responsible for delayed water excretion.^10,25

The concurrent use of beta-blockers with thiazide diuretics has shown a potential increase in the association of causing thiazide-induced hyponatraemia in this study. This association has not been reported in any other studies. However, it is plausible that beta-blockers may cause hyponatraemia as a result of their pharmacological properties. Beta-blockers inhibit the release of renin from the kidney, which leads to hypoaldosteronism, in turn enhancing renal loss of sodium, contributing to hyponatraemia.^27,28 Therefore, the concurrent use of beta blocker with a thiazide could have led to additive effects that resulted in hyponatraemia.

This study has several limitations. Firstly, the retrospective nature of this study hampered a complete retrieval of parameters that were reported in other studies, such as weight, creatinine clearance, physical immobility, duration of thiazide use and elderly home institutionalisation, water intake and dietary intake of sodium.^10,25 Secondly, as mentioned earlier in this paper, our study showed that low serum creatinine is a possible predictor of hospitalisation due to thiazide-induced hyponatraemia, contrary to other studies.^16,25,26 This is possibly due to selection bias in this study. The majority of the cases were community-dwelling adults not managed by SGH; hence, we cannot establish the overall health status. In contrast, our controls were selected from a known SGH patient population managed in the outpatient setting, with more chronic conditions. Hence, matching the number of chronic conditions and medications may possibly mitigate the selection bias.

Lastly, the sample size in our study may be too small to detect significant predictors. Moreover, given the small sample size, the results from our study may not be generalisable to the Singapore population. More research is warranted to determine the predictors of thiazide-induced hyponatraemia in older patients. Nevertheless, our study has found several factors that are associated with thiazide-induced hyponatraemia. The findings would be useful in planning for larger, controlled studies to evaluate the predictors of hospitalisation due to thiazide-induced hyponatraemia.

Conclusion

Our study suggested that low potassium levels and concurrent use of beta-blockers as potential predictors of hospitalisation due to thiazide-induced hyponatraemia in older Singaporeans. The awareness of the predictors for hospitalisation is crucial when prescribing, to promote the safe and effective use of thiazides in older adults.

Footnotes

Declaration of conflicting interest

None declared.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

James

Oparil

Carter

. 2014 evidence-based guideline for the management of high blood pressure in adults: Report from the panel members appointed to the Eighth Joint National Committee (JNC 8). JAMA 2014; 311: 507–520.

Krause

Lovibond

Caulfield

. Management of hypertension: Summary of NICE guidance. Br Med J 2011; 343: d4891.

Chobanian

Bakris

Black

. The seventh report of the joint national committee on prevention, detection, evaluation, and treatment of high blood pressure: The JNC 7 report. JAMA 2003; 289: 2560–2571.

Insua

Sacks

Lau

. Drug treatment of hypertension in the elderly: A meta-analysis. Ann Intern Med 1994; 121: 355–362.

Mulrow

Cornell

Herrera

. Hypertension in the elderly: Implications and generalizability of randomized trials. JAMA 1994; 272: 1932–1938.

Greenberg

Diuretic complications. Am J Med Sci 2000; 319: 10.

Clayton

Rodgers

Blakey

. Thiazide diuretic prescription and electrolyte abnormalities in primary care. Br J Clin Pharmacol 2006; 61: 87–95.

Byatt

Millard

Levin

Diuretics and electrolyte disturbances in 1000 consecutive geriatric admissions. J R Soc Med 1990; 83: 704–708.

Sunderam

Mankikar

Hyponatraemia in the elderly. Age Ageing 1983; 12: 77–80.

10.

Mann

SJ.

The silent epidemic of thiazide-induced hyponatremia. J Clin Hypertens 2008; 10: 477–484.

11.

Hwang

Kim

G-H.

Thiazide-induced hyponatremia. Electrolyte Blood Press 2010; 8: 51–57.

12.

Sonnenblick

Friedlander

Rosin

AJ.

Diuretic-induced severe hyponatremia. Review and analysis of 129 reported patients. Chest 1993; 103: 601–606.

13.

Zuern

Sauer-Schulz

Schnauder

[Hyponatremia–should it be ignored or diagnosed?–Case 8/2011]. Deutsche medizinische Wochenschrift (1946) 2011; 136: 1727.

14.

Routledge

O’Mahony

Woodhouse

Adverse drug reactions in elderly patients. Br J Clin Pharmacol 2004; 57: 121–126.

15.

Hamilton

Gallagher

Ryan

. Potentially inappropriate medications defined by STOPP criteria and the risk of adverse drug events in older hospitalized patients. Arch Internal Med 2011; 171: 1013–1019.

16.

Chow

Szeto

Wong

T-H

. Risk factors for thiazide-induced hyponatraemia. QJM 2003; 96: 911–917.

17.

Friedman

Shadel

Halkin

. Thiazide-induced hyponatremia: Reproducibility by single dose rechallenge and an analysis of pathogenesis. Ann Intern Med 1989; 110: 24–30.

18.

DeFronzo

Arieff

. Hyponatremia. In: Arieff

DeFronzo

(eds) Fluid, electrolyte and acid-base disorders. New York: Churchill Livingstone, 1995, pp.255–303.

19.

Sahai

Khurshid

Statistics in epidemiology: Methods, techniques and applications. Boca Raton, FL: CRC Press, 1995.

20.

Movig

Leufkens

Lenderink

. Association between antidepressant drug use and hyponatraemia: A case-control study. Br J Clin Pharmacol 2002; 53: 363–369.

21.

Kongkaew

Noyce

Ashcroft

DM.

Hospital admissions associated with adverse drug reactions: A systematic review of prospective observational studies. Ann Pharmacother 2008; 42: 1017–1025.

22.

Shea

Hammill

Curtis

. Medical costs of abnormal serum sodium levels. J Am Soc Nephrol 2008; 19: 764–770.

23.

Callahan

Caplan

. Economic impact of hyponatremia in hospitalized patients: A retrospective cohort study. Postgrad Med 2009; 121: 186–191.

24.

Rastogi

Pelter

Deamer

RL.

Evaluations of hospitalizations associated with thiazide-associated hyponatremia. J Clin Hypertens 2012; 14: 158–164.

25.

Clark

Shannon

Rosa

. Increased susceptibility to thiazide-induced hyponatremia in the elderly. J Am Soc Nephrol 1994; 5: 1106–1111.

26.

Barber

McKeever

McDowell

. A systematic review and meta-analysis of thiazide-induced hyponatraemia: Time to reconsider electrolyte monitoring regimens after thiazide initiation? Brit J Clin Pharmacol 2015; 79: 566–577.

27.

Bühler

Burkart

Lütold

. Antihypertensive beta blocking action as related to renin and age: A pharmacologic tool to identify pathogenetic mechanisms in essential hypertension. Am J Cardiol 1975; 36: 653–669.

28.

Sajadieh

Binici

Mouridsen

. Mild hyponatremia carries a poor prognosis in community subjects. Am J Med 2009; 122: 679–686.

A retrospective case-control study evaluating thiazide-induced hyponatraemia-related hospitalisation among older Singaporeans