Читать книгу Disorders of Fluid and Electrolyte Metabolism - Группа авторов - Страница 46

Since epidemiologic landmarks of hyponatremia are dependent on many factors, including the definition of hyponatremia used or the patient population examined, published reports on prevalence and incidence vary extensively. Nevertheless, hyponatremia undoubtedly constitutes the most common disorder of fluid and electrolyte balance throughout the world. In 1968, Owen and Campbell [14] observed that the mean serum [Na⁺] concentration of hospitalized patients was 5–6 mEq/L lower than in patients seen in the outpatient department. Anderson et al. [15] performed the first prospective analysis of patients with a serum [Na⁺] <130 mEq/L in a hospital in Denver (USA) in 1985 and reported a daily incidence and prevalence of 0.97 and 2.48%, respectively, throughout the various medical and surgical departments. Approximately, one-third of these patients were found to be hyponatremic already at the initial presentation [15]. However, these numbers likely underestimated the true rates since as many as 50% of patients at risk for hyponatremia did not have their serum [Na⁺] checked. Moreover, the dramatic changes in medical care since then, especially the widespread use of newly invented and highly effective drugs, very likely will have an impact on hyponatremia frequency.

Table 1. Etiology of hypotonic hyponatremia

More recently, in a large retrospective analysis spanning a 2-year period, 42.6% of more than 43,000 hospitalized patients in a 1,200-bed acute care hospital in Singapore developed hyponatremia (defined as serum [Na⁺] <136 mEq/L) at some point of their stay, 66% of those (i.e., 28.2% in total) were admitted already with hyponatremia. Of note, if the threshold for hyponatremia was lowered by only 1–135 mEq/L, the prevalence of hyponatremia on presentation dropped to 22.1%. The overall prevalence for more profound hyponatremia with serum [Na⁺] <126 mEq/L and <116 were 6.2 and 1.2%, respectively. In this Asian population, age was identified to be a strong independent risk factor while gender only had a minor influence. In contrast to the high prevalence of hyponatremia in hospitalized patients, the numbers were considerably lower with 7.2, 0.14, and 0.03% for serum [Na⁺] <136, <126, and <116 mEq/L, respectively, in 3 community-based primary care polyclinics that were also analyzed in the same study [16].

Waikar et al. [17] analyzed the database of 2 hospitals in Boston (USA) and identified a total of 98,411 eligible individuals that were admitted to either hospital over a 3-year interval. In 19.7% of those patients a serum [Na⁺] <135 mEq/L was reported during hospitalization; 14.5% were found to be hyponatremic on admission. In the latter group, hyponatremia was mild (i.e., serum [Na⁺] ≥130 mEq/L) in 12%, moderate (i.e., serum [Na⁺] 120–129 mEq/L) in 2.3%, and severe (i.e., serum [Na⁺] <120 mEq/L) in only 0.2% [17]. In another retrospective study, also conducted in a Boston academic medical center, more than 53,000 hospitalizations between 2000 and 2007 were examined. Primarily designed to assess association between serum [Na⁺] and mortality, hyponatremia was defined as a serum [Na⁺] <138 mEq/L since a significantly increased death rate was noticed below this threshold. Using this definition, hyponatremia was observed in 58% of all hospitalizations; 38% were community-acquired (i.e., hyponatremia was present on admission) and 20% were hospital-acquired. Of note, 5.7% of community-acquired hyponatremia aggravated during the hospital stay. While age and comorbidity score were identified as risk factors in both populations, more patients with community-acquired hyponatremia were treated in medical service and more patients with hospital-acquired hyponatremia were treated in surgical services [18].

In a prospective cohort study, Hoorn et al. [19] followed up on all hyponatremic patients over a 3-month period in a large university hospital in Rotterdam (The Netherlands). Hyponatremia, defined as a serum [Na⁺] ≤135 mEq/L, was found in 30% of 2,907 patients in who a serum [Na⁺] was obtained; severe hyponatremia ([Na⁺] <125 mEq/L) was present in 2.6%. In 51% of patients with severe hyponatremia, the condition was present at admission. The average serum [Na⁺] on admission in the group of severe hospital-acquired hyponatremia was 133 mEq/L, indicating that mild hyponatremia was already existing on admission at least in some patients. According to the authors’ interpretation, aggravation and deterioration of hyponatremia was then facilitated by factors introduced in the hospital rather than a de novo development of hyponatremia. The incidence of hyponatremia was significantly higher in internal medicine, surgery, and intensive care compared to all other departments [19]. The latest epidemiologic analysis of hyponatremia looking at 154,378 patients over a 5-year interval in a teaching hospital in Beijing (China) observed a prevalence of 17.5% for serum [Na⁺] <135 mEq/L with 13% for mild ([Na⁺] ≥130 mEq/L), 4.2% for moderate ([Na⁺] 120–129 mEq/L), and 0.3% for severe hyponatremia ([Na⁺] <120 mEq/L). Hyponatremic patients were significantly older and more frequently males than patients with normal serum [Na⁺] and there were significant differences between disease groups with infections, cancer, and cardiovascular diseases showing the highest hyponatremia rate [20]. In contrast to these studies, Zilberberg analyzed a database collating discharge and laboratory information from 39 hospitals over 2 years focusing on hyponatremia at the time of admission. To improve specificity, hyponatremia was only considered to be present when 2 serum [Na⁺] values <135 mEq/L within 24 h after admission were reported. Of almost 200,000 patients that were included in the analysis, only 5.5% were identified as being hyponatremic [21]. However, a more likely explanation might be rapid autocorrection in a high number of only mild hyponatremia cases. In addition, assessment of serum [Na⁺] is subject to biological and analytical variations, which can lead to a deviation of up to 3–4 mEq/L between 2 samples collected from the same patient at the same time [22]. Thus, despite the repeatedly shown statistical association between hyponatremia and morbidity and mortality, this measurement variability should be kept in mind when interpreting the observed high frequencies of mild and often asymptomatic hyponatremia. Another confounder with regard to measured sodium levels reflecting hypo-osmolality is the lack of assessments excluding pseudohyponatremia and normo- or hyper-osmolar states in some studies. Most [17–20] but not all studies, corrected sodium levels for hyperglycemia prior to analysis; Waikar et al. [17] reported that this mathematical correction reduced the frequency of hyponatremia by approximately 2% as compared to the raw data set.

Epidemiologic data on hyponatremia in the general population are sparse. In the Rotterdam study, between 1990 and 1993 all inhabitants of a suburb of Rotterdam aged ≥55 years (10,275 persons) were invited to participate in a prospective survey. Data on serum [Na⁺] were available for 5,179 individuals. The overall prevalence of hyponatremia (serum [Na⁺] ≤135 mEq/L) was 7.7%, with a mean serum [Na⁺] of 133 mEq/L [23]. In contrast, Mohan et al. [24] analyzed data from 14,804 adult patients in the National Health and Nutrition Examination Survey (NHANES) cohort between 1999 and 2004 and found a considerably lower weighed prevalence of 1.7%. However, there was no age restriction in NHANES, and the mean age in this cohort was markedly lower than in the Rotterdam study (45 vs. 70 years). Moreover, hyponatremia in NHANES was defined as serum [Na⁺] ≤133 mEq/L at least for the years 1999–2002. Recently, a subsequent analysis of data from NHANES comprising the time period between 2003 and 2006 revealed a more comparable hyponatremia prevalence of 5.6% [25]. In all studies, prevalence rose with higher age but only in the NHANES cohorts gender was identified as a significant risk factor, with women being more often affected. Gisby et al. [26] analyzed data from Swedish national population-based health registers and identified patients who were admitted to an inpatient service with a documented diagnosis of hyponatremia or SIADH (using the ICD-10 coding system) between 2001 and 2011. The prevalence of hyponatremia in that time period was 4.4%; the annual incidence was 14.8/100,000 in 2001 and increased steadily to 48.6/100,000 in 2011 [26]. Whether this development reflects a true increase in disease rate or rather an improved disease awareness is speculative. In this study, old age and female sex were again associated with a higher hyponatremia prevalence.