Clorox Concentration Technique for the Demonstration of Acid Fast by kby12992


									Clorox Concentration Technique for the Demonstration of
Acid Fast Bacilli in the Sputum
Romulo Uy, M.D.,** Charles Yu, M.D.,** Mario Juco, M.D.,** Cora Adlawan, M.D.,** Grace
Ruiz, M.D.,** Melecia Velmonte, M.D.** and Calixto Zaldivar, M.D.**

(1st Prize, PSMID- Hoechst Annual Research Contest, 1987. UP-PGH Medical Center)


            In countries like the Philippines where the prevalence of tuberculosis isi still high and here culture facilities
are oftentimes unavailable, the identification of the tubercle bacilli in the sputum by microscopy is still the most
accepted method for screening and follow-up of TB patients. Efforts therefore have to be geared toward improving
yield of the bacilli in the sputum.
            In the Philippine General Hospital, the recovery of the acid-fast bacilli in the sputum was relatively low and
concentration methods have been utilized to improve yield. This study was undertaken to compare the yield of TB
bacilli by the usual direct method against two concentration techniques, which used sodium hypochlorite (Clorox) and
            A total of 297 sputum specimens were collected from 101 subjects with clinical and radiologic evidence
highly suggestive of pulmonary tuberculosis. The mean age of these patients was 43.6 with a range from 17-76 years.
The results showed that the direct smear had a positive yield of 35.4% while the concentration methods revealed a
higher percentage, 45.8% for clorox and 46.1% for sputufluol. The difference was of statistical significance with a p
value of < 0.001; there was a lesser degree of concordance between the two methods. In addition, there was no
statistical significance between the two concentration methods, clorox versus sputufluol, with a p value of ≥ 0.05. The
degree of concordance was high between the two methods.
            Thus, concentration methods were superior to the direct method in the recovery of the tubercle bacilli in the
sputum. Clorox likewise can be a suitable alternative to sputufluol, an imported agent used in this hospital. In
epidemiological case finding activities, it can replace the direct method to increase the yield of AFB in the smear. [Phil
J Microbiol Infect Dis 1988; 17(1):13-18]

Key Words: sodium hypochlorite, TB, AFB smear, tuberculosis


         Tuberculosis (TB) persists as an important health problem particularly in developing
countries like the Philippines. According to the WHO figures, this country has the highest
mortality rate worldwide 1 (Figure 1). Based on the 1987 survey of the Department of Health,
although there was a decline in the mortality and morbidity of TB, it is the third leading cause of
death in this country with a rate of 55 per 100,000 and ranks fifth in the leading causes of
morbidity with a rate of 233.12 (Figure 2).In countries where the prevalence rate for tuberculosis
is high, the sputum stained smear for acid-fast bacilli is the major epidemiological tool in the
diagnosis of pulmonary tuberculosis. Smear positive tuberculous patients are given the highest
priority in case finding and treatment since these patients are the most dangerous sources of
infection. 3 The direct smear for AFB can detect one acid-fast bacillus per field when a milliliter
of sputum contains at least 1,000,000 tubercle bacilli assuming an even distribution of acid-fast
bacilli in the specimen and smear.3 Thus, it would be better to concentrate the organisms in the
         As early as 1915, Petroff 4 reported that Mycobacterium tuberculosis survived exposure
to 4 percent sodium hydroxide for 24 hours, whereas other organisms were killed. In the next
decade, newer agents were introduced. Krasnow and Wayne 4 reported that 3% sodium hydroxide
was actually toxic to tubercle bacilli and recommended the use of trisodium phosphate. Krasnow
and associates introduced zephiran into the trisodium phosphate, the former as a major
decontaminating agent and the latter serving to liquefy sputum. 5 Other techniques utilized the
phenol-autoclave concentration,6 N-acetyl-L-cysteine-sodium hydroxide (NALC-NaOH) method
of Kubica and associates,7 and Clorox concentration method by Oliver and Reuser.8
         Clorox contains 5.25% sodium hypochlorite as its active ingredient.8 Oliver and Reusser
reported that it can concentrate more TB bacilli in the smear compared to other methods by
digesting bulky sediments, pus, blood and mucus leaving a sediment consisting almost entirely of
tubercle bacilli. These studies8,9 were done several years ago without statistical analysis. Since
then, there has not been much interest regarding this reagent as a diagnostic tool for TB.

                                                                   Figure 2. Tuberculosis in the Philippines

   Figure 1: Tuberculosis mortality per 100,000

        In the Philippine General Hospital the rate of recovery of acid-fast bacilli in thesputum is
low (Table 1). The direct smear positivity rate (total positive smear/total number of specimens
ranged from 4.5 to 10.6% (with a mean of 7.3%). For TB bacilli culture the PGH laboratory since
1985 has been using a special decontaminating agent imported from Germany (Sputufluol), the
active ingredient of which is essentially sodium hypochlorite. The concentration of sodium
hypochlorite in this reagent is relatively low (about 0.02%) compared to that of the local
commercial version clorox (about 4.5%).These reagents were assayed in the Pharmacology
Department of the University of the Philippines. Medline search since 1966 revealed no
randomized well controlled trial had been done to confirm the 1942 report of Oliver and Reusser
using clorox for concentration of TB bacilli sputum smears.

Table 1. PGH data on direct smear and culture (1980-1986)

                                                   Culture   Results
                                             Positive            Negative             Total
Direct Smear                              No.         %        No.       %        No         %
Positive                                  204      (62.2)      124     (37.8)     328       (7.3)
Negative                                   32       (0.8)     4155     (99.2)    4187      (92.7)
Total                                     236       (5.2)     4279     (94.8)    4515     (100.0)
Source: Unpublished data by the authors

       This pilot study compares the yield of TB bacilli by the usual direct smear method
against the two concentration methods employing sodium hypochlorite manufactured
commercially as clorox, a locally available solution, and sputufluol, a reagent imported from


         A total of 101 patients were admitted to the study from November 1986 to July 1987.
Patients with clinical and radiologic evidence highly suggestive .of active pulmonary tuberculosis
and without previous intake of anti-TB medications for the past 3 months were included in the
trial. On entry to the study, these patients were provided with sterile wide-mouthed .bottles and
instructed on the proper way of collecting their sputum. Three consecutive early morning sputum
specimens were collected amounting was to at least 10 ml each.10 Each sputum specimen was
processed by three methods: the direct, clorox and sputufluol techniques (Figure3).
         A direct smear of each specimen collected was done immediately and the remain der was
divided equally into two 50-ml screw-capped culture tubes. The tubes were labeled A and B
respectively; tube A was used for the clorox mixture and Tube B for sputufluol method.
         For the direct smears a nickel-chrome wire loop was dipped into the sputum from the
mixture in the wide-mouthed bottle. A smear was made onto a chemically clean slide. The most
purulent, bloody and mucoid portion of the specimen was selected as recommended by different

                    Figure 3. Schematic diagram of the procedure of sputum specimen

        For the clorox concentration method the tube labeled A was placed on top of a serologic
shaking machine and was shaken for 15 to 20 minutes. Clorox was added in a 1:1 dilution to the
mixture after which it was centrifuged for 10 minutes at 3000 rpm. The supernatant liquid was
discarded and the sediment was smeared onto a new, chemically clean slide by means of a nickel-
chrome wire loop. For the sputufluol concentration method, the same procedure was done but
sputufluol instead of clorox was used for Tube B.
        The sediments were spread over an area of 200 square mm (10 x 20 mm) and stained
using the method of Kinyoun. 11
        For cultures, the remaining sediments were neutralized with 1N HCI using
phenolphthalein as indicator (solution turned colorless after the addition of the acid) and cultured
onto Lowenstein-Jensen medium using the drop method. All the tubes were incubated at37°C and
examined weekly for eight weeks for growth before it was finally discarded.
        All the slides were read by a qualified medical technologist who had no access to any
information that might result in bias (another medical technologist prepared the slides). The
reader reported the smears as positive or negative for acid-fast bacilli. Positive smears were
further classified based on the American Thoracic Society scale 10,11,12 (Figure 4). A quantitative
number was also specified by the reader based on the schema shown in Figure 4. As per
recommendation by different authors, all slides were read under oil immersion with 3
1ongitudinal sweeps equivalent to about 300 fields. A minimum of 3 acid-fast bacilli per slide
was required for the smear before it was considered as positive.

Figure 4. Method of reporting of acid-fast bacilli in the smear as recommended by the American Thoracic

Number of organisms seen                                      Report
1. 0 in entire slide                                          No AFB seen
2. 1-2 in entire slide                                        Repeat collection
3. 3-9 per slide                                              Rare
4. 10 or more per slide                                       Few
5. More than one per OIF                                      Numerous

        Chest x-rays were read officially by a senior radiologist and classified as to minimal,
moderately advanced or far-advanced and whether a cavity was present or not.13 Patients with
previous intake of anti-tuberculosis drugs for the past three months and those without cough were
excluded from the study.
        The results were analyzed by the chi square test for paired samples specifically the Stuart
Maxwell statistics at alpha = 0.05 to compare the direct smear versus the clorox method, direct
versus sputufluol and clorox versus sputufluol techniques with respect to the number of acid-fast
bacilli detected based on the American Thoracic Society criteria mentioned above. Kappa
agreement and its 95% confidence interval were computed to determine further the agreement
between the three methods.


        A total of 297 sputum specimens were collected from 101 patients with a mean age of
43.5 years (17-76 years). There were 66 males and 35 females with a 1.9:1.0 ratio. The clinical
characteristics of these patients are shown in Table 2.

Table 2. Clinical profile of patients included in the study

                                        Number                                    Percentage
1. Cough                                101                                       100.0
2. Weight loss                          62                                        61.4
3. Fever                                53                                        53.5
4. Anorexia                             49                                        49.0
5. Hemoptysis                           45                                        45.0

        The radiologic findings of these patients are shown in Table 3. Of the population under
study, cavitary lesions in the chest x-ray were seen in 53 patients or 52.5% of 101 patients. Far
advanced lesions comprised 45.5%, moderately advanced in 49.5% and minimal active lesions in
4.9% of all cases. All of the cases included in this study were classified as Class III based on the
American thoracic Society classification of tuberculosis.14
Table 3. Radiologic interpretation of the chest x-rays

                                             Number                          Percentage
1. Far advanced with cavity                  34                              33.7
2. Fat advanced without cavity               12                              11.9
3. Moderately advanced with cavity           19                              18.8
4. Moderately advanced without cavity        31                              30.7
5. Minimal active                            5                               4.9

         The results of the microscopic examination of the different methods are summarized in
Table 4. The direct smear microscopy yielded 64.3% (191/297) negative smears compared to the
clorox-treated and sputufluol-treated sputum specimens with 54.2% (161/297) and 53.9%
(160/297) negative smears respectively (p <0.001).

Table 4. Results of microscopic examination in 297 sputum specimen prepared by different smear methods

                                                Microscopic Examination Results*
                         Direct Smear                     Clorox treated                   Sputufluol treated
Results            No.                %               No.                %              No.                  %
Negative           191               64.3             161               54.2            160                 53.9
Rare                 9                3.0               7                2.4              7                  2.4
Few                 28                9.4              16                5.4             19                  6.4
Numerous            69               23.3             113               38.0            111                 37.4
Total              297              100.0             297              100.0            297                100.0
p value ( 0.001)

          Positive results were noted in only 106 specimens of direct smears while more specimens
yielded positive results when concentrated, 136 and 137 positive smears respectively for clorox
and sputufluol. Most of these positive results were noted and recorded as numerous in the two
concentrated methods than the direct smear, 37.38% compared to 23.3 % (p <0.001).
          When the results were lumped as either positive or negative alone, the percentage of
positivity of the direct smear was 35.7% (106/297); that of clorox and sputufluol were 45.8% and
46.1% respectively, (p <0.001).
          Table 5 matched the results of the different methods utilized. The direct smear yielded
more negative results compared to both concentration techniques. Concentration of the sputa
increased the positivity rates of the direct method. An additional 33 more specimens which were
negative with the direct smear were found positive with the clorox technique and 35 more with
statistically significant results at p < 0.001. However, there were 3 clorox-treated specimens and 4
sputufluol-treated specimens that were negative but turned out positive with the direct smear
method (Table 6).

Table 5. Matched data between direct and concentration methods

                                                Clorox (Sputufluol)*
                         Negative                               Positive
Direct Smear                                  Rare               Few               Numerous              Total
Negative              158 (156)               5 (4)              6 (7)              22 (24)               191
    Rare                 1 (3)                1 (1)              4 (2)               3 (3)                 9
    Few                  1 (1)                0 (2)              4 (5)              23 (20)               28
    Numerous             1 (0)                                   2 (5)              65 (64)               69
Total                 161 (160)               6 (7)             16 (19)            113(111)              297
*The numbers enclosed in the parenthesis refer to sputufluol
Table 6. Correlation between direct and concentration methods

                                                  Clorox (Sputufluol)*
                                       Negative                          Positive                      Total
Negative                          158           (156)               33               (35)               191
Positive                            3            (4)               103              (102)               106
Total                             161           (160)              136              (137)               297
*The numbers in parenthesis refer to sputufluol

        From the above results 2 positive results were observed more in specimens which were
concentrated with either clorox or sputufluol than .the usual direct smear (Table 6).
        The kappa statistics for agreement beyond chance alone at 95% confidence interval
between the direct smear and clorox was 0.59 and between direct and sputufluol, 0.57. There was
a high degree of discordance between direct and both concentration methods/
        Comparable results were obtained when clorox and sputufluol were matched as seen in
Table 7. There was no difference between the total number of negative results (160 against 161)
and the cumulative number of positive results (136 versus 137). At 95% confidence interval, the
kappa statistics of agreement beyond chance alone between the two methods was 0.80, showing a
high degree of concordance between the two concentration methods.

Table 7. Matched data between clorox and sputufluol methods*

                        Negative                                 Positive
                                               Rare                Few                      Numerous       Total
Negative                    152                  2                   2                         5            161
Rare                         3                   2                   1                         1              7
Few                          2                   2                   9                         3             16
Numerous                     3                   1                   7                        102           113
Total                       160                  7                  18                        111           297
*There was no statistical difference between these two methods (p value > 0.05)

       Culture was positive only in 16.8% (17/101) of all the study patients. The growth of
Mycobacterium tuberculosis was observed only in the specimens subjected to sputufluol. There
was no growth in those treated with clorox. Incidentally there were three patients with atypical
mycobacteria noted in culture. The species were not identified.


        A case of pulmonary tuberculosis is referred to by the World Health Organization as a
patient with bacteriologically confirmed disease.17 For underdeveloped countries and in remote
areas where culture facilities are unavailable, a bacteriologically confirmed disease is that in
which a patient is found positive with acid-fast bacilli in the sputum by microscopy. The smear
can establish a presumptive diagnosis of TB.
        The chances of finding the acid-fast bacilli in a smear increase with the concentration of
the bacilli in the specimen. Several authors4,7,8 utilized different concentration techniques to
increase the yield of positive acid-fast bacilli in the sputum and other body fluids, Table 8. It was
generally recommended that addition of an alkali to the sputa can digest fibrous tissues, blood
cells and other extraneous materials leaving a sediment, after centrifugation, consisting entirely of
the tubercle bacilli. 7,8,9,15
        Most of these studies were, however, time consuming and expensive. No statistical
analysis was done in most of these studies to confirm their findings.
        This study utilized sodium hypochlorite, which is locally available as clorox and an
imported one sputufluol and compared the results with the usual direct method. As a preliminary
report this study showed that concentration methods were far superior to the direct smear in the
recovery of acid-fast bacilli in the sputum. There were more positive results with both
concentration methods than the direct smear method (p < 0.001). Conversely, the latter method
yielded more negative results than the former two methods (p > 0.001).

Table 8. Different concentration methods used in smears

Authors                              Year                     Agents Used
Petroff4                             1915                     4% sodium hydroxide
Oliver and Reusser8                  1942                     Clorox (Sodium hypochlorite)
Muller and Chermock6                 1945                     Phenol autoclave
Krasnow and Wayne4                   1946                     Trisodium phosphate
Krasnow et al5                       1965                     Trisodium phosphate with Zephiran
Kubica and Kohn7                     1963                     N-acetyl-L-cysteine with sodium hydroxide

         A local study by Tech16 compared the direct method and the concentration method using
sodium hydroxide. The results showed that the direct smear method screened more positive
results than the sodium hydroxide. This s      tudy however was uncontrolled with no statistical
analysis done. In another major institution, a study in 197417 reported a 41.8% positivity rate
among far-advanced cases of tuberculosis using the direct method. Foreign literature has cited
different posit ivity rates from 30 to as high as 88% utilizing fluorescent microscopy. 15,18,19 Filho
and Fonseca20 reported a 36.1% positivity rate using sodium hypochlorite with centrifugation and
a 44.3% rate upon the addition of xylol as a flotation agent. Their results were comparable to this
present study. This study was however uncontrolled.
         The degree of agreement between the direct and concentration methods was also
computed. There was a lesser degree of concordance between the direct and both the
concentration techniques. However, when both the concentration methods were compared, the
degree of concordance between clorox and sputufluol was high at a kappa value of 0.80. From
these preliminary results, one can use either of the two concentration methods in the laboratory.
However, considering the cost and availability of the agents, clorox may substitute for the
imported sputufluol.
         Clorox is a relatively cheap solution and is readily available in the rural areas. It is also
noted that indeed this reagent is bactericidal considering the absence of growth in culture of all
the specimens treated with it. This can be advantageous because in laboratories without biologic
safety cabinets, this method can minimize infection among laboratory personnel. Thus, the
material is safe for handling.
         In the Philippine General Hospital, there is ongoing use of this agent, clorox, but at a
lower concentration (0.02%) in the recovery of the tubercle bacilli in culture.
Another study will have to be undertaken this time to improve the yield in culture of the tubercle
bacilli in this hospital as part of its thrust to function as a level III as well as level I
mycobacteriology laboratory. The authors would like to suggest this method to the Department of
Health as an alternative to the direct method in epidemiological case-finding activities and
thereby complement the department's goal o t eradicate tuberculosis by the year 2000. 2 This must
however be on the background of good technical expertise and adequate laboratory facilities in
the peripheral health centers in the country. In this connection the authors, in cooperation with the
Department of Health, will be conducting an anti-TB control project in an urban barangay in San
Andres Bukid with the implementation of this method in case-finding activities in the said
community and in monitoring the patient's response and progress to treatment.

         In summary, a total of 297 sputa were collected from 101 subjects .with clinical and
radiological evidence of pulmonary tuberculosis. Mean age of these patients was.43.6 with ages
ranging from 17-76 years o1d. The results showed that the positivity rate of the direct smear was
35.7%; that of clorox was 45.8% and 46.1% for sputufluol. Statistical significance was noted
between the direct method and the two concentration methods as to the recovery of acid-fast
bacilli in the sputum p < 0.001 There was also a great degree of discordance between these two
methods (direct versus concentration). In addition, there was no statistical significance noted
when the two concentration methods clorox against sputufluol were compared (p. > 0.05). Clorox
can therefore be a. useful substitute to the more expensive sputufluol in this hospital to increase
the yield of AFB in the sputum smear and should replace the direct method in epidemiological
ease finding activities.


          The authors would like to extend their sincere gratitude to the following people who helped in the completion
of this project: Undersecretary Manuel Roxas, who supplied us with anti-TB medications, Mr. Eliseo Tura and Mrs.
Medy Episcope our medical technologist and Miss Cynthia Cordero, the statistician of the project.


1.    WHO/Japan International Tuberculosis Course: Tuberculosis Control. The Research Institute of Tuberculosis,
      JapanAntituberculosisAssociation, April 1979.
2.    Department of Health, Manila, Philippines. National TuberculosisPrevalenceSurvey,1987
3.    Tomas K. Tuberculosis case finding and chemotherapy, World Health Organization, 1979:7
4.    Krasnow I, Wayne L. Sputum digestion: The mortality rate of tubercle bacilli in various digestion systems. The Am J ClinPathol
      1966; 45(3):352-355.
5.    Krasnow I, Kidd GC. The effect of a buffer wash of sputum sediments digested with zephiran trisodium phosphate on the
      recovery of acid-fast bacilli. Am J Clin Pathol 1965; 44:238-240.
6.    Muller HE, Chermock RL. A rapid staining technique for acid-fast organisms. J Lab Clin Med 1945; 30:169-171.
7.    Kubica GP, Kaufmann AJ, Dye WE. Comments on the use of the new mucolytic agents; N-Acetyl-L-Cysteine, as a s            putum
      digestant for the isolation of mycobacteria. Am Rev Respir Dis 1964; 89:284-286,
8.    Oliver J, Reusser TR. Rapid method for the concentration of tubercle bacilli. Am Rev Respir Dis 1942; 45:450-452.
9.    Tarshis MS, Lewis WG. Use of clorox and tri-sodium phosphate in demonstration of acid-fast bacilli in sputum. Am J Clin
      Pathol 1949;19:688.
10.   David HL. Bacteriology of mycobacteriosis. DHEW Public No. (CDC) 76-8316. Center for Disease Control, Atlanta, Georgia.
11.   Kubica GP, DavidHL. The mycobacterium. In: Gradwohls Clinical Laboratory Methods and Diagnosis. AC Sonnenwith, L Jarett
      (eds) 8th Edition. CV Mosby Co. 1980.
12.   Wayne LG. Microbiology of tuberculosis. In: Pulmonary Diseases and Disorders AP.Fishman (ed) McGraw Hill Book
      Company, 1980.
13.   Paul and Juhl's Essentials of Roentgen Interpretation. Harper and Row, 4 th edition, 1981.
14.   American Thoracic Society. Diagnostic Standards and Classification of Tuberculosis and other Mycobactetial Diseases (14th
      edition), November, 1980.
15.   Kubica GP, Gross WM, Hawkins JE, .Sommers HM, Vestal AL, Wayne LG. Laboratory services for mycobacterial diseases.
      Am Rev Respir Dis 1975; 112: 775.
16.   Tech, DM. A comparative study of the smear methods in the identification of acid-fast bacilli from the sputum. JPMA 1965;
17.   Cruz BV, Morales CT, Manalo FM. Sputum microscopy in an emergency TB hospital: A study of the results in 2021 male cases
      in the Quezon Institute (1974-1976). Chest Dis 1976.
18.   Boyd JC, Mart J. Decreasing reliability of acid-fast smear: Techniques for detection of tuberculosis . Ann Intern Med
19.   Strumpf J, Tsang A, Sayre JW. Re-evaluation of sputum staining for the diagnosis of pulmonary tuberculo sis. Am Rev Respir
      Dis 1979;119: 599-602.
20.   Filho CP, Fonseca KS. The flotation method for detection of tubercle bacilli in sputum smears. Tubercle 1979; 60:105-107.

To top