Serum chemistry abnormalities in adult patients with seizures

ORIGINAL CONTRIBUTION seizures, adult, chemical abnormalities

Serum Chemistry Abnormalities in Adult Patients with Seizures To determine whether checking serum chemistries is a useful procedure, records were reviewed to identify patients presenting to a university hospital emergency department (ED) with seizures. In a six-month period, 112 adult patients made 126 visits for evaluation and treatment of recent seizures. Analyzing each visit individually, 11I patients (88.1%) had serum electrolytes determined; 96 (76.2%) had a serum calcium level measured; and 92 (73%) had a serum magnesium level measured. The overall incidence of seizures due primarily to derangements in serum chemistry was 2.4%, incIuddng two hypoglycemic seizures and one seizure related to hemodialysis. Severe aberrations in serum chemistries occurred most frequently in alcoholics; nonalcoholics had relatively few abnormal test results unless they had preexisting renal failure or diabetes mellitus. In ED patients, the incidence of seizures due primarily to derangements in serum chemistry is very low. Grouping of patients by medical history can help direct test ordering by identifying those at risk for abnormalities in serum chemistry [Powers RD: Serum chemistry abnormalities in adult patients with seizures. Ann Emerg Med May 1985;14:416-420.] INTRODUCTION There are numerous etiologies for seizures, and a complete diagnostic workup of each patient would require sophisticated testing that is beyond the scope of most emergency departments (EDs). T h e proper method of assessment of an emergency patient with a seizure is unclear, but measurement of serum electrolytes often is included as an initial step. Metabolic and electrolyte disturbances account for 10% to 15% of isolated seizures in hospitalized patients, I but whether a search for this etiology is useful in patients presenting to an ED for evaluation of a recent seizure is unknown. This study used a retrospective chart review to determine the type and extent of serum chemistry abnormalities present in ED patients with seizures.

Robert D Powers, MD Charlottesville, Virginia From the Emergency Medical Services and the Department of Internal Medicine, University of Virginia Medical Center, Charlottesville, Virginia. Received for publication June 12, 1984. Revision received September 7, 1984. Accepted for publication December 6, 1984. Presented at the University Association for Emergency Medicine Annual Meeting in Louisville, Kentucky, May 1984. Address for reprints: Robert D Powers, MD, Box 523, University of Virginia Hospital, Charlottesville, Virginia 22908.

METHODS Nurses and housestaff in the ED of the University of Virginia Hospital routinely obtained blood for serum electrolytes, BUN, and creatinine (SMA-7) on virtually all patients presenting with a seizure. Serum levels of magnesium and calcium frequently were measured as well. Patients were identified by examining the log book of the University of Virginia ED from January through June 1982. An attempt was made to review the ED record of all aduk patients whose admitting or discharge diagnosis from the ED contained the word seizure. Pediatric patients (those under age 14) and patients referred from other hospitals were excluded. Of 188 patients who met these criteria, charts of 181 (96.2%) were available for review. Of these, 126 patient visits in which the discharge diagnosis from the ED was "seizure" or "probable/possible seizure" were identified. The remaining 55 patients were judged to have had events other than seizures; most frequently these were diagnosed as syncopal episodes or anxiety at~ tacks. ED records of the 126 patient visits for seizures were examined for pertinent information, including demographic data, medical history, and laboratory values determined as a part of the ED patient evaluation. Statistical analysis was done using Fischer's exact test.

14:5May 1985

Annals of Emergency Medicine

416/63

SERUM CHEMISTRY ABNORMALITIES Powers

RESULTS Patient Demographics The 126 adult visits for seizures over a six-month period represented 0.7% of total ED registrations during that time. These visits were made by 112 different patients, who ranged in age from 15 to 85 years (average age, 41.8 years). Sixty-five patients (.58%) were white, 46 (41.1%) were black, and one (1%) was Oriental. Seventy-seven (68.8%)were men. Forty-four (39.3%) were identified in their charts as ethanol abusers, and 71 (63.4%) had a previously diagnosed seizure disorder.

Results of ED Laboratory Tests For purposes of analysis, each patient visit to the ED was considered a separate encounter, so that 126 sets of patient data were identified. A multiple chemistry panel (SMA-6) was determined for 111 (88.1%). Ninety-six (76.2%) had a serum calcium level measured and 92 (73%) had a serum magnesium level measured. The individual values of the serum chemistries determined are summarized (Table 1), and the degree to which the abnormalities deviated from the normal range is shown (Figure 1). Abnormal values were arbitrarily considered to be potentially significant if they were more than two standard deviations outside the limits of the reference range of the University of Virginia clinical laboratories. No reference values were available for nonfasting glucose; therefore, a potentially significant range was defined arbitrarily as being outside the limits of 50 to 250 mg/dL. All patients with high creatinine levels had k n o w n c h r o n i c r e n a l failure; both hypoglycemic patients had previous episodes of low blood sugar; and significant hyperglycemia occurred only in patients with poorly controlled diabetes mellitus. There was a very strong correlation between low levels of magnesium and alcoholism. Magnesium levels less tl~an or equal to 1.5 mEq/L were significantly more prevalent in alcoholics (12 of 40) than in nonalcoholics (1 of 52) (P = .00006). No seizures were specifically attributed to hypomagnesemia, although eight of 18 patients (44.4%) with alcohol-related seizures had a serum magnesium less than or equal to 1.5 mEq/L. 64/417

TABLE 1. Results of serum chemistry testing of emergency department patients with seizures

Sodium (136-145)¢ Potassium (3.5-5.0) Chloride (96-106) Bicarbonate (24-30) BUN (10-26) Creatinine (0.7-1.5) Glucose (60-150) Magnesium (1.8-2.8) Calcium (8.5-10.5) *All measurements are which are mg/dL. 1-Normal ranges.

N

Range*

Below Normal # (%)

Above Normal # (%)

111

130-153

6 (5.4)

6 (5.4)

111

2.8-5.9

20 (18.0)

5 (4.5)

111

88-113

8 (7.1)

23 (20.7)

111

8-37

43 (38.7)

5 (4.5)

111

1-63

34 (30.6)

5 (4.5)

111

0.4-14.0

6 (5.4)

5 (4.5)

111

27-832

2 (1.8)

23 (20.7)

31 (33.6)

1(1.1)

92

0.7-3.2

96 7.7-11.1 5 (5.2) mEq/L except BUN, creatinine, and glucose,

Relationship Between Test Results and Medical History Patients were separated into three subgroups according to historical criteria. Group 1 included patients who had never had a seizure prior to presentation. A summary of their abnormal test results is shown (Figure 2). In this group only one seizure was attributed to an abnormal serum chemistry value; this was a hypoglycemic seizure in a diabetic caused by accidental insulin overdose. Nonalcoholic patients with histories of seizures were designated Group 2; their s e r u m c h e m i s t r y abnormalities are shown (Figure 3). Two seizures in this group were attributed to derangements in serum chemistrg, including a hypoglycemic seizure and a dialysis-related disequilibrium seizure. In both cases, the underlying disorders causing the metabolic disturbances had been identified previously, and each patient had suffered prior seizures of identical etiology. Group 3 consisted of patients identified in their charts as chronic ethaAnnals of Emergency Medicine

2 (2.1)

nol abusers, regardless of whethei they presented with a first seizure or with recurrent seizures. Despite the overlap with Group 1, it was believed that alcoholics presented an easily identified subgroup of patients at high risk for electrolyte disorders regardless of seizure history, and that they should be considered separately from others. As a subgroup, alcoholics had the largest number of abnormal chemistry results (Pigure 4). Severe derangements occurred primarily in levels of bicarbonate and magnesium. No seizures in this group were attributed specifically to disorders of serum chemistry, although most patients were given parenteral magnesium supplementation because of the presumed contribution of hypomagnesemia to the initiation of alcohol withdrawal seizures. 2

Etiology of Seizures Patients were categorized (Table 2) according to diagnosis on discharge from the ED. Eighty-three of the patients (65.4%} were discharged to their

14:5 May 1985

%>2SD Above Normal Range

:I

Na

K

CI

HC03 BUN Creat Mg

GLUC

Fig. 2. Group 1: Patients without previous seizures.

w

%>2SD Below Normal Range

Ca

T

I0 15 20

55 patient visits (44% of total) No abnormal sodium, potassium, chloride, or calcium 1 patient with glucose < 60 mg/dL 2 patients with BUN

> 27 mg/dL

3 patients with Mg + + < 1.5 mEq/L

homes and 43 (34.6%) were admitted to the hospital. The overall incidence of seizures attributed specifically to serum c h e m i s t r y abnormalities was 2.4% {3 of 126), including the two hypoglycemic seizures and the single seizure due to dialysis-related electrolyte disequilibrium.

DISCUSSION These data provide little justification for r o u t i n e d e t e r m i n a t i o n of serum chemistry on all adult ED patients with seizures. The incidence of seizures in ED patients due primarily to disorders of serum c h e m i s t r y is quite low, and the data suggest that these patients can be identified by historical criteria, minimizing the need for random screening laboratory tests. The absence of unanticipated abnormalities in serum chemistry fails to support the contention that such testing may be useful to uncover other medical problems, even if it cannot be justified as a means of identifying the cause of the seizure that prompted the ED visit. Previous studies3, 4 have examined the utility of determining serum dec14:5May 1985

Fig. 1. Incidence of potentially significant serum chemistry abnormalities in adult emergency department patients with seizures.

t r o l y t e levels in ED p a t i e n t s w i t h first-time seizures; the tests were of diagnostic value in less than 5% of patients. The present study extends this observation to all adult ED seizure patients, and it suggests that these abnormalities can be anticipated if the p a t i e n t ' s history, m e d i c a t i o n , and clinical presentation are taken into account. It is not surprising that a low serum bicarbonate was the m o s t c o m m o n abnormality detected in this series, because m a n y patients who have recently suffered a grand mal seizure will have a moderate acidosis. Thus the low bicarbonate supports the recent occurrence of a seizure, but it does not help to identify its cause. Although hyponatremia is a known cause of seizures, no patient in this series had a h y p o n a t r e m i c seizure. Nonetheless, hyponatremia probably should be sought in all patients with first seizures, as well as those with historical features and physical findings that suggest its presence. Serum levels of chloride and potassium do not relate directly to the etiology or management of seizures, but they will Annals of Emergency Medicine

be available f r e q u e n t l y because of their inclusion in multiple electrolyte panels. As such, occasional patients who need potassium repletion will be identified, but it is likely that appropriate historical predicting factors would be present in the majority of these individuals. This study does not support the usefulness of serum glucose as a rand o m l a b o r a t o r y test. Both p a t i e n t s with hypoglycemic seizures had dearcut historical reasons for low blood glucose. In the absence of insulin therapy, liver disease, or h y p o g l y c e m i c agent ingestion, spontaneous hypoglycemia to a level that could cause seizures is rare; a routine screen for it is not likely to be productive, s Clinicians wishing to know blood glucose levels can easily substitute a bedside determination using Dextrostix ® or Glucometer ® rather than the more expensive and time-consuming laboratory methods. The renal function parameters BUN and creatinine provided no useful information about the patients. When present, severe renal impairment was predicted by historical factors; the only seizure attributed to renal failure o c c u r r e d in a p a t i e n t on dialysis whose evaluation logically would have included determination of BUN and creatinine. Routine testing of laboratory values not included in the electrolyte panels is even more difficult to justify. Hypomagnesemia has such a strong association with alcoholism that its presence is a virtual certainty in alcoholic ED patients, whether they are seen for seizures or other reasons. 6 The ED practice of giving alcoholics parenteral m a g n e s i u m s u p p l e m e n t a t i o n as a matter of course limits the clinical value of concurrent determination of serum levels, as these are a poor ref l e c t i o n of t o t a l b o d y m a g n e s i u m stores. 7 Significant hypomagnesemia is extremely unusual in nonalcoholics who otherwise are in good health; because of its association with seizure disorders, however, the magnesium levels 418/65

SERUM CHEMISTRY ABNORMALITIES Powers

Fig. 3. Group 2: Previous seizure disorder, nonalcoholic.

48 patient visits (38% of total)

Fig. 4. Group 3: Alcoholics, w i t h or without previous seizures.

No potentially significant abnormalities in sodium, potassium, chloride, calcium, or magnesium 8 patients with bicarbonate < 20 mEq/L

should be measured in patients with first-time seizures or poorly controlled seizures despite adequate anticonvulsant m e d i c a t i o n levels. This study does not support the routine measurem e n t of m a g n e s i u m levels in nonalcoholics if t h e y have a well-described seizure disorder that has been evaluated previously. Of all the serum chemistry values determined in our study, the calcium provided the least useful diagnostic or therapeutic information. Only 7% of the levels were outside the n o r m a l range; the few high ones were easily attributed to preexisting renal failure, and the low ones were of dubious significance because the level of serum albumin in these patients was not obtained. The determination of serum calcium in seizure patients could be restricted to those with first-time seizures, where hypocalcemia has been reported as a cause, 8 and to those occasional patients with multiple endocrine or medical problems that suggest the presence of abnormalities in calcium homeostasis. Grouping seizure patients by history allows a more rational approach to appropriate laboratory testing. All seizure patients can be categorized as first-seizure patients, nonalcoholics with a known seizure disorder, or alcoholics. Although this study did not specifically support it, standard neurology 9 and e m e r g e n c y m e d i c i n e lo texts recommend a thorough laboratory evaluation of a patient presenting with a first-time seizure. Even though the yield is low, determination of serum chemistries is indicated in seizure patients in the initial attempt to identify a specific etiology for the seizure. Patients whose medical history is unknown or unobtainable must be treated as first-time seizure patients. Nonalcoholics with a k n o w n seizure disorder are u n l i k e l y to have spontaneous severe electrolyte or renal function abnormalities. Routine testing of these individuals can be deferred in the absence of historical factors or physical findings to the contrary. T h e p a t i e n t w i t h idiopathic epilepsy who has a typical seizure is

3

1 patient with glucose

< 60 mg/dL

2 patients with BUN

> 30 mg/dL

1 patient with creatinine

> 2.0 mg/dL

51 patient visits (40% of total) Subgroup with largest number of abnormal lab results 11 patients with bicarbonate < 20 mEq/L 12 patients with Mg+ +

probably best managed by determining anticonvulsant levels and adjusting therapy accordingly, rather than by embarking on a lengthy search for m u c h less likely causes. Alcoholics have a high incidence of serum electrolyte and renal function abnormalities (Figure 4), and their ED e v a l u a t i o n for a s e i z u r e p r o b a b l y should include an electrolyte panel. Even when patients have a history of alcohol withdrawal seizures, the risk of related or unrelated severe metabolic disturbances is high enough to justify a routine electrolyte screen. Serum magnesium can be expected to be low in alcoholics, but because the actual level is useful for determining replacement therapy, it may be reasonable to measure magnesium levels in alcoholic patients with seizures. Routine determination of serum calcium in alcoholics is not justified by this study, although it should be done if the seizure was the patient's first.

SUMMARY Abnormalities in serum chemistry occur with variable frequency in ED patients with seizures; the overall incidence of seizures due specifically to these abnormalities is low. In most cases, severe electrolyte disturbances can be predicted from such historical factors as diabetes, renal failure, and alcoholism. In the absence of these or other clinical indicators of metabolic imbalance, routine determination of serum chemistries in ED patients with seizures is unlikely to provide information that is useful for diagnosis or therapy. Prospective studies axe needed to determine whether

< 1.5 mEq/L

TABLE 2. Etiology of seizures - rill patients

% Patients

Etiology Anticonvulsant noncompliance

25%

Known seizure d i s o r d e r typical seizure

25%

New-onset seizure

20%

ETOH-related seizure

16%

"Probable" seizure - - ? etiology

14%

Hypoglycemic seizure

2%

limited use of serum chemistry studies in the ED evaluation of seizure patients would result in significant savings of t i m e and m o n e y w i t h o u t compromising patient care. The author thanks Chuck Byam, EMT, and Don Kaiser, DrPH, for data processing and statistical assistance; Joseph Chance, MD, for manuscript review; and Susan Loving for technical and secretarial support.

REFERENCES 1. Bauer G, Niedermeyer E: A c u t e convulsions. Clin Electroencephalogr 1979j 10:127-144.

2. Brown CG: The alcohol withdrawal syndrome. Ann Emerg Med 1982;11:276280. 3. Rosenthal RM, Helm ML, Waeckerle JF: First-time major seizures in an emergency department. A n n Emerg Med }

66/419

Annals of Emergency Medicine

14:5 May 1985

1980;9:242-245. 4. Russo L8, Goldstein KH: The diagnostic assessment of single seizures. Arch Neurol 1983;49:744-746. 5. Ensinck JW, Williams RH: Disorders causing hypoglycemia, in Williams RH (ed}: Textbook of Endocrinology, ed 6. Philadelphia, WB Saunders Co, 1981.

6. Graber TW, Yee AS, Baker FJ: Magnesium: Physiology, clinical disorders, and therapy. Ann Emerg Med 1981;10:49-57. 7. Massey SG: Managing disorders of serum magnesium concentration. Drug Therapy (Hosp) Feb 1980:43-52. 8. Cox RE: Hypoparathyroidism: An unusual cause of seizures. Ann Emerg Med 1983;12:314-315.

9. Goldensohn ES, Glaser GH, Goldberg MA: Epilepsy, in Rowland LP (ed): Mettit's Textbook of Neurology. Philadelphia, Lea and Febiger, 1984. 10. Tomlanovich MC, Yee AS: Seizure, in Rosen P, Baker FJ, Braen GR, et al (eds): Emergency Medicine: Concepts and Clinical Practice. St Louis, CV Mosby Co, 1983, p 1339-1358.

1985 ACEP Council Resolution Deadline The ACEP Council will meet September 7-8, 1985, in Las Vegas, Nevada. All proposed amendments to the Constitution and Bylaws of the American College of Emergency Physicians must be received by the Council Secretary no later than 90 days in advance of the annual meeting. For the 1985 meeting that date is June 10. All other resolutions should be submitted to the Secretary no later than 45 days in advance of the meeting. This year that date is July 24. Address resolutions to: Colin C Rorrie, Jr, PhD, SecretaG ACEP Council, PO Box 619911, Dallas, Texas 75261-9911. Advance submission of resolutions is preferable to floor resolutions because it permits chapter review and allows staff time to prepare analyses of fiscal impact and previous College action. 14:5 May 1985

Annals of Emergency Medicine

420/67