Serum Chloride Analysis, Bromide Detection, and the Diagnosis of Bromism R O B E R T E. W E N K , M.D., J A C K A. L U S T G A R T E N , P H . D . , N. J O H N P A P P A S , R O B E R T I. LEVY, M.D.,
AND R O G E R J A C K S O N ,
M.D.,
M.D.
Departments of Pathology and Medicine, Sinai Hospital of Baltimore and Crownsville Hospital Center, Baltimore, Maryland
ABSTRACT
BROMIDE THERAPY has largely been out- tion, which in turn depends on the method moded by the use of other hypnotics and of chloride analysis, and 3) the interference sedatives. Bromide poisoning has therefore of bromide with various methods of become an unfamiliar disorder; a 1975 chloride analysis. review of 100 toxicologic agents 1 that omits bromide from the list attests to the obMethods solescence of bromism as a suspected diagnosis. Nevertheless, occasional cases of Effect of Inorganic Bromide on Methods bromide intoxication (bromism) are still of Chloride Analysis reported. 3,10 Since some cases are first sus(1) Aqueous Br~ solutions (100 mmol. pected as a result of apparently elevated Br~ per 1.) were prepared by weighing serum chloride values, diagnosis is partly out sodium bromide salt. Dilutions yielda function of the chloride method utilized ing solutions of concentrations ranging by clinical laboratories. 2,9 We have evalfrom 1 to 100 mmol. per 1. were made. uated 1) the occurrence of bromide in These solutions, which contained no contemporary serum samples of hospitalchloride, were subjected to chloride analyized patients, 2) the dependence of the sis by five common procedures: ion-selecdiagnosis of bromism on bromide detective electrode (Stat/Ion, Technicon Corp., Received February 18, 1975; accepted for publica- Tarrytown, N.Y.), coulometry (Chlorition April 10, 1975. dometer, American Instrument Co., Silver Address reprint requests to Dr. Wenk: Department of Pathology, Sinai Hospital, Baltimore, Spring, Md.), and thiocyanate-colorimetry Maryland 21215. (Chloride Analyzer, I n s t r u m e n t a t i o n 49
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
Wenk, Robert E., Lustgarten, Jack A., Pappas, N. John, Levy, Robert I., and Jackson, Roger: Serum chloride analysis, bromide detection, and the diagnosis of bromism. Am J Clin Pathol 64: 4 9 - 5 7 , 1976. Current methods for determining serum chloride concentration vary in specificity. Laboratory detection of bromide in serum is usually dependent on spuriously high chloride values. In the absence of historical information, the chemical identification of bromide in serum may be the only diagnostic sign of bromism. Bromide may be found in as many as 1% of sera obtained from inpatients of a general hospital. Ion-selective electrodes and some thiocyanate methods for chloride analysis can be highly sensitive to bromide (interference). Chloride electrodes are also sensitive to iodide. (Key words: Chloride; Bromide; Bromism; Ion-selective electrode [ISE].)
50
WENK ET
AL.
AJ.C.P.—Vol.
65
were compared with the apparent measured CI - concentrations in all three kinds of solutions. Comparisons were expressed as a ratio of apparent (measured) chloride minus known chloride divided by known bromide: [Apparent Cl~] - [known CI - ] / [known B r - ]
40
60
80
100
FIG. 1. Effects of bromide and bromide plus chloride to constant halide of 100 mmol. per 1. in aqueous solution on various methods for chloride. a, DuPont ACA plus chloride; b, DuPont ACA— no chloride; c, Stat/Ion—no chloride; d, Stat/Ion, lus chloride; e, Chloride Analyzer plus chloride; / Chloride Analyzer, no chloride; g, Chloridometer, no chloride; h, Chloridometer plus chloride; i, Theoretical one-to-one response of bromide to chloride.
Laboratories, Lexington, Mass.; Automatic Clinical Analyzer, DuPont, Wilmington, Del.; Sequential Multiple Analyzer 6/60, Technicon Corp.).* (2) Similar solutions were prepared in protein-based media and were analyzed by the same procedures. (3) Bromide solutions of similar concentrations were also prepared in saline solution so that total halide, bromide plus chloride, equalled a constant 100 mmol. per 1. T h e concentrations of the two halides together were considered more clinically relevant since physiologically, bromide replaces chloride 12 and because these halides occur together in laboratory samples. These solutions were analyzed for chloride by the procedures noted above. (4) The known concentrations of B r * Apparent chloride test results have not been corrected for those values exceeding the linearity of any particular method.
Frequency of Bromide Occurrence in Serum Samples (1) Random sampling was made of hospitalized patients' sera over several months. The samples, 133 sera, were analyzed for
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
20
BROMIDE CONCENTRATION - mmol / liter
Ratios were plotted graphically against the known bromide concentrations for each method in order to delineate bromide interference with chloride analyses. (5) A study of ion-selective electrodes (ISE) was made because of the relatively recent emergence of ion-selective electrode technology as a routine clinical laboratory tool. To study the effect of loss of selectivity over time, electrode responses using a brand-new CI - sensor electrode and one utilized for several months in our laboratory were compared. Study was made of interelectrode variation, fatigue, change in response time, and the effects of "poisoning" by B r - , I - , and F - . T h e sensors were exposed to pure bromide solutions, to pure chloride solutions, or to mixtures of B r - and CI - in solutions of constant halide concentration. Exposure of the chloride electrodes to bromide (10 mmol. B r - + 90 mmol. Cl per 1.), for periods of 1.5 minutes and for 1 hour, was followed by tests for sensor damage, i.e., refractory or augmented response to aqueous CI - solutions of 100 mmol. per 1. and solutions of CI - + B r (total halide = 100 mmol. per 1., Br~ = 5, 10, 20 mmol. per 1.). The sensors were similarly exposed to aqueous iodide solutions (1 mmol. I + 99 mmol. CI - per 1.) and fluoride solutions (1 mmol. F - + 99 mmol. CI - per 1.).
January 1976
51
SERUM Cl~ METHODS, BROMIDE AND BROM1SM
Results Method Studies (1) Aqueous bromide solutions, analyzed for chloride by several methods, registered variable apparent chloride concentrations (Fig. 1). All methods showed positive interference—the higher the bromide concentration, the higher the ap-
5.0
A (no chloride)
4.0
3.0
-
o ex CO
~
2.0
01
1.0
-
id *e
t=*=
3.0
2.0
1.0
20
40
60
80
100
BROMIDE CONCFNTRATION - mmol / liter
FlG. 2. Effects of bromide (A) and bromide plus chloride (B) to constant halide of 100 mmol. per 1. aqueous solution on: a, DuPont ACA; b, Stat/ Ion; c, SMA 6/60; d, IL Chloride Analyzer; e, Chloridometer.
parent chloride concentration (Figs. 2A and 4). T h e ISE results followed a sigmoidal response curve, while the curves with other methods were linear or curvilinear in response to increasing bromide concentrations. (2) Solutions of bromide in the presence of chloride, with final halide totalling 100 mmol. per 1., when analyzed for chloride, showed similar responses (Figs. 1 and 2B). The electrode response was diminished in magnitude and appeared curvilinear rather than sigmoidal. Other methods showed response curves that were either parallel to, or identical to, the curves derived from pure bromide solutions (coulometry, Chloride Analyzer thiocyanate, ACA thiocyanate). (3) Results of chloride analysis of pooled
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
bromide spectrophotometrically. 13 T h e bromide method was evaluated by calculating the percentage of bromide recovered from aqueous solutions (2-100 mmol. per 1.) in the presence of various concentrations of chloride ( 0 - 9 8 mmol. per 1.). Recovery was also determined from serum pools containing similar concentrations of halides. Precision was determined from statistical analysis of 18 repeated B r - determinations on serum pools containing 1.8 to 3.7 mmol. per 1. (2) It was found that routine analysis of electrolytes by ion-selective electrodes over six months occasionally produced spuriously high chloride values (range 110-306 mmol. per 1.). These values were not accompanied by cation abnormalities and caused a decrease in the normal anion gap ( [ N a + ] minus the [Cl~ + H C 0 3 - ] ) . Spurious chloride elevations occurred in about 0.75% of samples. 9 Each anomalous result initiated a search for historical evidence that the patient used bromide agents. The serum in question was chemically analyzed for bromide. (3) During the course of study, three cases of bromism were diagnosed, and are described. (4) An attempt was made to determine the effect of small amounts of bromide in serum on chloride analysis. One of the authors intentionally ingested 900 mg NaBr (about 14.5 mg. per kg.). Chloride analysis by coulometry, electrode, and thiocyanate (Chloride Analyzer) methods was carried out.
52
WENK ET
FlG. 3. Analysis of pooled sera containing various bromide concentrations with constant chloride, by various chloride methods, a, DuPont ACA; b, Stat/ Ion; c, 1L Chloride Analyzer; d, Chloridometer.
serum in the presence of constant chloride of 97 mmol. per 1. containing bromide over the range of 1-20 mmol. per 1. are shown in Figure 3. The ISE method yielded, at low bromide concentrations, an a p p a r e n t ratio (chloride/bromide) greater than 4, which decreased to approximately 3 over the Br~ range of 10-20 mmol. per 1. For the DuPont ACA, this ratio was also high at low bromide concentrations; however, the ratio decreased to about half that for the ISE method at high B r - concentrations. Other methods also showed initially high ratios, which, however, were approximately 1 or slightly less over the same bromide concentration range. (4) The magnitude of inappropriate response of each chloride measurement to bromide interference is expressed more fully in Figures 2 and 3. In the absence of chloride (Fig. 2A) the Cl~ electrode method is less selective at low concentrations of bromide and more selective at high concentrations. The reverse is true for the ACA thiocyanate procedure. (Some of the curves are foreshortened at low levels since the methods are known not to be linear at these extreme, non-
A.J.C.P.—Vol.
65
physiologic concentrations. Manufacturer's recommendations were followed when possible.) Least interference among the methods was noted for the IL thiocyanate procedure, SMA 6/60 method, and coulometry. The coulometric method showed slightly more interference at the lowest bromide levels. Chloride analysis of solutions containing chloride as well as bromide showed similar, but less augmented, interference. Again, the electrode and ACA procedures showed greater responses to bromide than to chloride on a mole-for-mole basis. Other methods signalled one mmol of chloride per liter for each mmol of bromide present. (5) Used and new chloride electrodes did not differ greatly in their responses to pure bromide and bromide plus chloride with constant halide concentrations (Table 1). The greatest difference was found at low concentration of bromide in the absence of chloride. Exposure of electrodes to bromide for short intervals (minutes) caused a refractory or slow response in the one sample cup that followed the bromide-containing solution (Fig. 5). Repeated exposure of an electrode to bromide at intervals of about ten minutes (every six samples on the turntable) appeared to decrease the magnitude of erroneous response to subsequent bromide solutions; however, no predictable magnitude of response was apparent. Exposures of electrodes to bromide, iodide, or fluoride solutions for an hour did not appear to affect subsequent chloride analysis-response further. However, iodide (1 mmol. per 1.) showed interference: a three- to tenfold response as apparent chloride (mmol. per 1.) was signalled for the one mmol of iodide present per liter. Exposure of an electrode to I - for one hour slightly increased the electrode response to bromide, but had no longterm effect. Response to subsequent chloride solutions was appropriate. Fluoride caused no interference either with short or long exposures.
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
BROMIDE CONCENTRATION - mmol / liter
AL.
January 1976
53
SERUM CI" METHODS, BROMIDE AND BROMISM
Patient Studies (1) Of the 133 patients' sera, selected at random over several weeks, two showed the presence of bromide (both approximately 1 mmol. per 1.) by the gold chloride method. This method is considered relatively specific for serum bromide determination. 13 O u r data on constant total halide solutions with bromide in concentrations ranging from 2 to 100 mmol. per 1. showed average recoveries of 100.3%, range 93.8 to 105.0%. Precision and recovery data on serum pools are shown in Table 2.
-
5.0 > APPARENT . CHLORIDE / BROMIDE X
3.0 5
140
TO
2.0 1
•
120
SZ ^F
° CD
1.0 o
100
' APPARENT CHLORIDE
< 20
0
20
40
60
80
100
INDIVIDUAL HALIDE CONCENTRATION - mmol / liter
FIG. 4. Electrode responses to aqueous solutions containing increasing chloride and bromide concentrations.
an initial chloride value of 216 mmol. per 1. Chloride analysis was repeated colorimetrically (IL Chloride Analyzer) and coulometrically (Aminco Chloridometer), yielding values of 106 and 108
Table 1. Responses of Old and New CI Electrodes to Bromide Solutions and to Bromide plus Chloride Solutions (at Constant Halide of 100 mmol. per 1.) New Electrode
Br" Concentration
Apparent Cl-/Br-
Apparent Cl"/Br (in Presence of CI")
10 20 30 40 50 60 70 80 90 100
5.6 5.0 3.7 2.9 2.4 2.1 1.9 1.7 1.6 1.6
2.9 2.5 2.2 2.0 1.9 1.7 1.6 1.5 1.5 1.4
Old Electrode _
CrVBr-
Apparent C I ' / B r (in Presence of CI")
4.0 4.1 3.5 3.1 2.5 2.1 1.9 1.7 1.7 1.6
3.0 2.5 2.2 2.0 1.9 1.8 1.6 1.6 1.5 1.5
Apparent
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
(2) Patients who had spurious chloride elevations in their sera are tabulated below along with historical information and serum B r - concentrations (Table 3). Some of these patients were receiving bromide by prescription; others had taken it in proprietary preparations; still others had no evidence of bromide administration. T h e bromide compounds determined by investigation are listed. (3) The three cases of bromism found in six months of study are summarized as follows: Case 1. R.B., a 78-year-old white woman, was admitted to Sinai Hospital with lethargy, disorientation to time, place, and person; and dehydration. Electrolyte studies using selective ion electrodes revealed
, 160
54
WENK ET AL. 20Br
UJ
oo t—
ELE
o
210 200
UJ
> •—
190
ELE
(J
180
o
170
mol / litei
(/z>
160 150
E
(J z
130 120
9X (_> 110 H-
z
UJ
< Q_
AND R O G E R J A C K S O N ,
M.D.,
M.D.
Departments of Pathology and Medicine, Sinai Hospital of Baltimore and Crownsville Hospital Center, Baltimore, Maryland
ABSTRACT
BROMIDE THERAPY has largely been out- tion, which in turn depends on the method moded by the use of other hypnotics and of chloride analysis, and 3) the interference sedatives. Bromide poisoning has therefore of bromide with various methods of become an unfamiliar disorder; a 1975 chloride analysis. review of 100 toxicologic agents 1 that omits bromide from the list attests to the obMethods solescence of bromism as a suspected diagnosis. Nevertheless, occasional cases of Effect of Inorganic Bromide on Methods bromide intoxication (bromism) are still of Chloride Analysis reported. 3,10 Since some cases are first sus(1) Aqueous Br~ solutions (100 mmol. pected as a result of apparently elevated Br~ per 1.) were prepared by weighing serum chloride values, diagnosis is partly out sodium bromide salt. Dilutions yielda function of the chloride method utilized ing solutions of concentrations ranging by clinical laboratories. 2,9 We have evalfrom 1 to 100 mmol. per 1. were made. uated 1) the occurrence of bromide in These solutions, which contained no contemporary serum samples of hospitalchloride, were subjected to chloride analyized patients, 2) the dependence of the sis by five common procedures: ion-selecdiagnosis of bromism on bromide detective electrode (Stat/Ion, Technicon Corp., Received February 18, 1975; accepted for publica- Tarrytown, N.Y.), coulometry (Chlorition April 10, 1975. dometer, American Instrument Co., Silver Address reprint requests to Dr. Wenk: Department of Pathology, Sinai Hospital, Baltimore, Spring, Md.), and thiocyanate-colorimetry Maryland 21215. (Chloride Analyzer, I n s t r u m e n t a t i o n 49
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
Wenk, Robert E., Lustgarten, Jack A., Pappas, N. John, Levy, Robert I., and Jackson, Roger: Serum chloride analysis, bromide detection, and the diagnosis of bromism. Am J Clin Pathol 64: 4 9 - 5 7 , 1976. Current methods for determining serum chloride concentration vary in specificity. Laboratory detection of bromide in serum is usually dependent on spuriously high chloride values. In the absence of historical information, the chemical identification of bromide in serum may be the only diagnostic sign of bromism. Bromide may be found in as many as 1% of sera obtained from inpatients of a general hospital. Ion-selective electrodes and some thiocyanate methods for chloride analysis can be highly sensitive to bromide (interference). Chloride electrodes are also sensitive to iodide. (Key words: Chloride; Bromide; Bromism; Ion-selective electrode [ISE].)
50
WENK ET
AL.
AJ.C.P.—Vol.
65
were compared with the apparent measured CI - concentrations in all three kinds of solutions. Comparisons were expressed as a ratio of apparent (measured) chloride minus known chloride divided by known bromide: [Apparent Cl~] - [known CI - ] / [known B r - ]
40
60
80
100
FIG. 1. Effects of bromide and bromide plus chloride to constant halide of 100 mmol. per 1. in aqueous solution on various methods for chloride. a, DuPont ACA plus chloride; b, DuPont ACA— no chloride; c, Stat/Ion—no chloride; d, Stat/Ion, lus chloride; e, Chloride Analyzer plus chloride; / Chloride Analyzer, no chloride; g, Chloridometer, no chloride; h, Chloridometer plus chloride; i, Theoretical one-to-one response of bromide to chloride.
Laboratories, Lexington, Mass.; Automatic Clinical Analyzer, DuPont, Wilmington, Del.; Sequential Multiple Analyzer 6/60, Technicon Corp.).* (2) Similar solutions were prepared in protein-based media and were analyzed by the same procedures. (3) Bromide solutions of similar concentrations were also prepared in saline solution so that total halide, bromide plus chloride, equalled a constant 100 mmol. per 1. T h e concentrations of the two halides together were considered more clinically relevant since physiologically, bromide replaces chloride 12 and because these halides occur together in laboratory samples. These solutions were analyzed for chloride by the procedures noted above. (4) The known concentrations of B r * Apparent chloride test results have not been corrected for those values exceeding the linearity of any particular method.
Frequency of Bromide Occurrence in Serum Samples (1) Random sampling was made of hospitalized patients' sera over several months. The samples, 133 sera, were analyzed for
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
20
BROMIDE CONCENTRATION - mmol / liter
Ratios were plotted graphically against the known bromide concentrations for each method in order to delineate bromide interference with chloride analyses. (5) A study of ion-selective electrodes (ISE) was made because of the relatively recent emergence of ion-selective electrode technology as a routine clinical laboratory tool. To study the effect of loss of selectivity over time, electrode responses using a brand-new CI - sensor electrode and one utilized for several months in our laboratory were compared. Study was made of interelectrode variation, fatigue, change in response time, and the effects of "poisoning" by B r - , I - , and F - . T h e sensors were exposed to pure bromide solutions, to pure chloride solutions, or to mixtures of B r - and CI - in solutions of constant halide concentration. Exposure of the chloride electrodes to bromide (10 mmol. B r - + 90 mmol. Cl per 1.), for periods of 1.5 minutes and for 1 hour, was followed by tests for sensor damage, i.e., refractory or augmented response to aqueous CI - solutions of 100 mmol. per 1. and solutions of CI - + B r (total halide = 100 mmol. per 1., Br~ = 5, 10, 20 mmol. per 1.). The sensors were similarly exposed to aqueous iodide solutions (1 mmol. I + 99 mmol. CI - per 1.) and fluoride solutions (1 mmol. F - + 99 mmol. CI - per 1.).
January 1976
51
SERUM Cl~ METHODS, BROMIDE AND BROM1SM
Results Method Studies (1) Aqueous bromide solutions, analyzed for chloride by several methods, registered variable apparent chloride concentrations (Fig. 1). All methods showed positive interference—the higher the bromide concentration, the higher the ap-
5.0
A (no chloride)
4.0
3.0
-
o ex CO
~
2.0
01
1.0
-
id *e
t=*=
3.0
2.0
1.0
20
40
60
80
100
BROMIDE CONCFNTRATION - mmol / liter
FlG. 2. Effects of bromide (A) and bromide plus chloride (B) to constant halide of 100 mmol. per 1. aqueous solution on: a, DuPont ACA; b, Stat/ Ion; c, SMA 6/60; d, IL Chloride Analyzer; e, Chloridometer.
parent chloride concentration (Figs. 2A and 4). T h e ISE results followed a sigmoidal response curve, while the curves with other methods were linear or curvilinear in response to increasing bromide concentrations. (2) Solutions of bromide in the presence of chloride, with final halide totalling 100 mmol. per 1., when analyzed for chloride, showed similar responses (Figs. 1 and 2B). The electrode response was diminished in magnitude and appeared curvilinear rather than sigmoidal. Other methods showed response curves that were either parallel to, or identical to, the curves derived from pure bromide solutions (coulometry, Chloride Analyzer thiocyanate, ACA thiocyanate). (3) Results of chloride analysis of pooled
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
bromide spectrophotometrically. 13 T h e bromide method was evaluated by calculating the percentage of bromide recovered from aqueous solutions (2-100 mmol. per 1.) in the presence of various concentrations of chloride ( 0 - 9 8 mmol. per 1.). Recovery was also determined from serum pools containing similar concentrations of halides. Precision was determined from statistical analysis of 18 repeated B r - determinations on serum pools containing 1.8 to 3.7 mmol. per 1. (2) It was found that routine analysis of electrolytes by ion-selective electrodes over six months occasionally produced spuriously high chloride values (range 110-306 mmol. per 1.). These values were not accompanied by cation abnormalities and caused a decrease in the normal anion gap ( [ N a + ] minus the [Cl~ + H C 0 3 - ] ) . Spurious chloride elevations occurred in about 0.75% of samples. 9 Each anomalous result initiated a search for historical evidence that the patient used bromide agents. The serum in question was chemically analyzed for bromide. (3) During the course of study, three cases of bromism were diagnosed, and are described. (4) An attempt was made to determine the effect of small amounts of bromide in serum on chloride analysis. One of the authors intentionally ingested 900 mg NaBr (about 14.5 mg. per kg.). Chloride analysis by coulometry, electrode, and thiocyanate (Chloride Analyzer) methods was carried out.
52
WENK ET
FlG. 3. Analysis of pooled sera containing various bromide concentrations with constant chloride, by various chloride methods, a, DuPont ACA; b, Stat/ Ion; c, 1L Chloride Analyzer; d, Chloridometer.
serum in the presence of constant chloride of 97 mmol. per 1. containing bromide over the range of 1-20 mmol. per 1. are shown in Figure 3. The ISE method yielded, at low bromide concentrations, an a p p a r e n t ratio (chloride/bromide) greater than 4, which decreased to approximately 3 over the Br~ range of 10-20 mmol. per 1. For the DuPont ACA, this ratio was also high at low bromide concentrations; however, the ratio decreased to about half that for the ISE method at high B r - concentrations. Other methods also showed initially high ratios, which, however, were approximately 1 or slightly less over the same bromide concentration range. (4) The magnitude of inappropriate response of each chloride measurement to bromide interference is expressed more fully in Figures 2 and 3. In the absence of chloride (Fig. 2A) the Cl~ electrode method is less selective at low concentrations of bromide and more selective at high concentrations. The reverse is true for the ACA thiocyanate procedure. (Some of the curves are foreshortened at low levels since the methods are known not to be linear at these extreme, non-
A.J.C.P.—Vol.
65
physiologic concentrations. Manufacturer's recommendations were followed when possible.) Least interference among the methods was noted for the IL thiocyanate procedure, SMA 6/60 method, and coulometry. The coulometric method showed slightly more interference at the lowest bromide levels. Chloride analysis of solutions containing chloride as well as bromide showed similar, but less augmented, interference. Again, the electrode and ACA procedures showed greater responses to bromide than to chloride on a mole-for-mole basis. Other methods signalled one mmol of chloride per liter for each mmol of bromide present. (5) Used and new chloride electrodes did not differ greatly in their responses to pure bromide and bromide plus chloride with constant halide concentrations (Table 1). The greatest difference was found at low concentration of bromide in the absence of chloride. Exposure of electrodes to bromide for short intervals (minutes) caused a refractory or slow response in the one sample cup that followed the bromide-containing solution (Fig. 5). Repeated exposure of an electrode to bromide at intervals of about ten minutes (every six samples on the turntable) appeared to decrease the magnitude of erroneous response to subsequent bromide solutions; however, no predictable magnitude of response was apparent. Exposures of electrodes to bromide, iodide, or fluoride solutions for an hour did not appear to affect subsequent chloride analysis-response further. However, iodide (1 mmol. per 1.) showed interference: a three- to tenfold response as apparent chloride (mmol. per 1.) was signalled for the one mmol of iodide present per liter. Exposure of an electrode to I - for one hour slightly increased the electrode response to bromide, but had no longterm effect. Response to subsequent chloride solutions was appropriate. Fluoride caused no interference either with short or long exposures.
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
BROMIDE CONCENTRATION - mmol / liter
AL.
January 1976
53
SERUM CI" METHODS, BROMIDE AND BROMISM
Patient Studies (1) Of the 133 patients' sera, selected at random over several weeks, two showed the presence of bromide (both approximately 1 mmol. per 1.) by the gold chloride method. This method is considered relatively specific for serum bromide determination. 13 O u r data on constant total halide solutions with bromide in concentrations ranging from 2 to 100 mmol. per 1. showed average recoveries of 100.3%, range 93.8 to 105.0%. Precision and recovery data on serum pools are shown in Table 2.
-
5.0 > APPARENT . CHLORIDE / BROMIDE X
3.0 5
140
TO
2.0 1
•
120
SZ ^F
° CD
1.0 o
100
' APPARENT CHLORIDE
< 20
0
20
40
60
80
100
INDIVIDUAL HALIDE CONCENTRATION - mmol / liter
FIG. 4. Electrode responses to aqueous solutions containing increasing chloride and bromide concentrations.
an initial chloride value of 216 mmol. per 1. Chloride analysis was repeated colorimetrically (IL Chloride Analyzer) and coulometrically (Aminco Chloridometer), yielding values of 106 and 108
Table 1. Responses of Old and New CI Electrodes to Bromide Solutions and to Bromide plus Chloride Solutions (at Constant Halide of 100 mmol. per 1.) New Electrode
Br" Concentration
Apparent Cl-/Br-
Apparent Cl"/Br (in Presence of CI")
10 20 30 40 50 60 70 80 90 100
5.6 5.0 3.7 2.9 2.4 2.1 1.9 1.7 1.6 1.6
2.9 2.5 2.2 2.0 1.9 1.7 1.6 1.5 1.5 1.4
Old Electrode _
CrVBr-
Apparent C I ' / B r (in Presence of CI")
4.0 4.1 3.5 3.1 2.5 2.1 1.9 1.7 1.7 1.6
3.0 2.5 2.2 2.0 1.9 1.8 1.6 1.6 1.5 1.5
Apparent
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
(2) Patients who had spurious chloride elevations in their sera are tabulated below along with historical information and serum B r - concentrations (Table 3). Some of these patients were receiving bromide by prescription; others had taken it in proprietary preparations; still others had no evidence of bromide administration. T h e bromide compounds determined by investigation are listed. (3) The three cases of bromism found in six months of study are summarized as follows: Case 1. R.B., a 78-year-old white woman, was admitted to Sinai Hospital with lethargy, disorientation to time, place, and person; and dehydration. Electrolyte studies using selective ion electrodes revealed
, 160
54
WENK ET AL. 20Br
UJ
oo t—
ELE
o
210 200
UJ
> •—
190
ELE
(J
180
o
170
mol / litei
(/z>
160 150
E
(J z
130 120
9X (_> 110 H-
z
UJ
< Q_
