STUDIES ON MICROORGANISMS IN SIMULATED ROOM ENVIRONMENTS II. THE SURVIVAL RATES OF STREPTOCOCCI IN THE DARK' LEON BUCHBINDER AND EARLE B. PHELPS DeLamar Institute of Public Health, College of Physicians and Surgeons, Columbia University, New York City Received for publication January 31, 1941 In the first paper in this series (Phelps and Buchbinder, 1941) it was shown that when a broth culture of bacteria is sprayed into the air of an experimental chamber, a certain fraction of the nuclei resulting from evaporation of the spray droplets and containing viable bacteria have measurable settling velocities. Under our experimental technique, settling proceeds at such a rate that about 20 per cent of those bacteria-laden particles capable of settling do so during each hour. This geometric rate of decrease results in a residual population in the air at the end of twenty-four hours of about one-half of one per cent of the initial number capable of settling. The experiments reported in the present paper were undertaken to determine the survival rates of that portion of the bacteria which settle to the floor. The preliminary partition of the sprayed material into suspended and deposited groups is primarily a matter of physical properties of the material itself, incidentally complicated by an existing death rate among the bacteria. An exact description of survival rates, however, must include a definite measurement of death rates as well as settling rates among those organisms which have settled. It is this latter phase of the subject with which the present paper deals. The studies reported are limited to three strains of streptococci; an alpha-hemolytic streptococcus (29B) originally isolated from Supported by a grant from the John B. Pierce Foundation. 345
3463LEON BUCHBINDER AND EARLE B. PHELPS a human throat, a beta-hemolytic streptococcus of Group A (330) isolated from a human throat and pathogenic for mice, and a Group B beta-hemolytic streptococcus (090) of human origin and highly virulent for mice. METHODS Open sterile petri dishes whose bottoms were completely covered with sterile filter paper were placed on the floor of the experimental chamber. The petri dishes were attached in sets of six or eight by strings leading to the small door of the room through which they could be removed. An 18-hour blood broth culture diluted 1:10 or 1:25 with broth was then sprayed into the experimental chamber for about five minutes. The sprayer and the experimental chamber have been described in the preceding paper. In preliminary tests, other surfaces to receive the organisms in addition to filter paper had been studied, including plain glass, vaseline, glucose, dextrine, etc. but none of them had been found to be satisfactory. This is not to imply that the paper plates were ideal. The colonies which grow do not represent the actual numbers of viable organisms settled as determined by comparative tests made with exposed blood agar plates. Nevertheless the results obtained by the paper plate method have been found to be consistent in a large number of tests and the method is apparently satisfactory for the determination of survival rates. Two distinct methods were used for determining survival rates. In the first, all the plates were removed from the experimental room a few minutes after spraying. Alternate plates which served as controls were immediately flooded with blood agar and incubated for two or three days at 3700., at which time the colonies were counted. The remainder of the plates were kept at laboratory temperature in the dark for 24 hours and were then treated in the same manner as were the controls. A 24- hour rate was obtained by repeated trials, the test plates being subject to the vagaries of the laboratory temperature and humidity. In the second method the-plates were allowed to remain in
MICROORGANISMS IN ROOM ENVIRONMENTS 347 the experimental room until ready for pouring, determined daily for a number of successive days. the rate being RESULTS In table 1 are given the data of the tests carried out with the alpha-hemolytic streptococcus (29B). The experiment of January 16, 1940 was made with 24 control plates, (initial) and plates exposed for 24 hours. There was an average of 300 or- TABLE 1 Survival in the dark for one day of a strain of alpha-hemolytic streptococcus (29B) settled from the air of a room INITIAL 24 Houns DATE, 1940 1PER CEN M.. Number of Number of Number of Numberof SURVIVALM.S.T. plate organisms plates organism days 1/16 24 300 184 61 1.4 1/22 23 395 24 158 40 0.8 1/24 24 272 24 135 50 1.0 1/26 24 433 24 116 27 0.5 1/30 24 360 24 110 31 0.5 2/1 24 295 24 175 59 1.3 2/7 220 142 65 1.6 2/16 596 382 64 1.6 2/18 24 113 24 78 69 1.9 2/23 264 110 43 0.8 Averages... 51 E4.7 1.1 1.13 ganisms per plate at the start and of 184 twenty-four hours later, a survival of 61 per cent. It is clearly indicated in all this work that the death rates under a wide variety of conditions are nearly if not exactly geometric. We find it consistent therefore to employ the mathematical technique of dealing with death and survival rates as if they were exactly geometric. Such geometric rates are identified by a velocity constant, k, the instantaneous or momentary proportionate rate of decrease. Corresponding to each such k value there is a "half life" value or "median survival time" which is much more readily explained and understood. It is the time required under any existing death
3488LEON BUIOEBINDER AND EARLE B. PHELPS rate for a diminution of the population to one-half its initial value. During the next equal time interval the reduction will be to onequarter, etc. A survival after 24 hours of 61 per cent implies a 50 per cent survival after 1.4 days. We will employ this half life period therefore as a measure of the death rate, and record it as m.s.t. (median survival time). The reliability of the data of table 1 may now be investigated. The chief cause of variation lies in the unequal distribution of bacteria upon the plates exposed in the chamber. Yet we wish to utilize the spraying technique because it so closely simulates the actual conditions in a dwelling place. Hence, rather large numbers of parallel plates have been employed in each test. From a considerable amount of material available in this and other series of experiments, all based on the same initial technique of spraying and settling, we have computed for each set of from 6 to 24 plates a mean and standard deviation. The standard deviation, due solely to random distribution of the settling bacteria among a series of plates, is known to be the square root of the mean value. Actually we have found among 80 such sets an average ratio of standard deviation to square root of mean of 2.25 instead of unity. This represents the additional variability inherent in the technique and the ratio, 2.25, is adopted as a basis for a further study of reliability. From it, by the usual procedure, we derive a value of 8.1 for the s.d. of the mean, 300; of 9.2 for the value 184, and of 3 for the survival ratio 61 per cent, and 0.13 for the half life 1.4 days. Similar computations for each line of table 1 gives mean standard deviations for the survivals and for the half life value of 6 and 10 per cent respectively. The actual standard deviations among themselves of the series of 10 experiments are about three times as great as these values, indicating that in repeating the experiments day after day, other extraneous variables enter. It so happens that the series of ten sets nearly compensates for this additional variability and gives a reliability factor to the mean value of the set of about 9 and per cent for the survival ratios and the m.s.t. respectively. Table 2 contains similar data of four experiments with a Group
MICROORGANISMS IN ROOM ENVIRONMENTS A, beta-hemolytic streptococcus (330). The survival rates are too high to justify any conclusions from single experiments because of the experimental variation. Taking the entire set of four, however, we have a survival ratio of 95 i 3 per cent and a m.s.t. of thirteen days, but owing to the geometric character of the variations, having a standard deviation range of 9 to 28 days. TABLE 2 Survival in the dark for one day of a strain of beta-hemolytic streptococcus [group A, type 9 (880)] settled from the air of a room DATE, 1940 Noum- Nu INITIAL 24 ous = Num- Number of BSUEVIVA M.S.T. ber of orgnisof ber of Plato orgnimsoransm _ 3/5 8 534 8 471 88 5.4 3/6 8 816 8 790 97 23.0 3/ 8 2 8 116 95 13.6 3/13 8 189 8 199 100 Total... 32 415 4:8* 32 394 418 95 =13 13(9-28) * With the use of a lexian ratio of 2.25. 349 Table 3 contains the data of a single experiment with still another strain, 090 (Group B). This was a continuous study up to 13 days to illustrate the geometric nature of the death rate. In the last column we show values of a true geometric series decreasing by per cent each day. The standard deviations of the actual counts are about 20 at the start down to about 10 at the end. Variations for the computed values are often larger than expected as sampling error, again due to other variable conditions during the test. The evidence for a general geometric course in the death curve however is good. The m.s.t. is 5.5 days. The experiments reported this far were made in the laboratory at room temperature without any attempt to control the latter, which was not far from 250. In order to investigate the effect of temperature on the dark death rate a series of experiments was carried out with strain 29B exposed in the incubator under ~~~~~~~~~~~~~~~~~days
350 LEON BUCHBINDER AND EARLE B. PHELPS TABLE 3 Survival in the dark for 18 days of a strain of a beta-hemolytic streptococcus [group B(090)] settled from the air of a room DAYS AFTER SEEDING 0 1 2 3 4 56 7 89 10 11 13 NUBER OF PLATE 6 NIUMBUB O Reoovered 961 945 735 725 814 588 518 429 334 193 382 272 2 284 ORGANOIS PUB PLATE Computed 1071 945 830 730 643 566 498 437 384 338 298 262 230 203 Survival rate, geometric, used for "computed" is 88 per cent in one day. timated m.s.t. is 5.5 days. TABLE 4 The effect of temperature on the survival rate of strain 29B in the dark TUMPURATURZ OF INCUBATOR 55 C. 50 C. a-c. I40C. Time of exposure in minutes 40 53 68 30 40 50 30 40 50 80 106 133 60so 100 Number of colonies on control plates*... 300 263 214 103 161 133 201 276 270 264 251 184 137 148 137 Number of colonies on test plates*... 4 71 35 62 98 5 161 175 169 187 0 94 139 138 3 Survival rate (per cent).. 41 27 16 60 61 94 80 64 63 71 48 51 100 93 90 M.s.t. (minutes).... 31 28 25 41 56 555t 93 61 75 161 100 136 cot774 783 M.s.t. (hours)... 0.47 1.1 2.2 13 * Average of eight plates. t This result discarded. controlled temperatures. The results of this series are shown in table 4. Between 25 and 40 degrees the mean survival time is apparently reduced from 26 to 13 hours. Between 40 and 45 Es-
MICROORGANISMS IN ROOM ENVIRONMENTS degrees there appears to be a rather abrupt critical point which is entirely too marked to be attributed to variability of the data. It may be significant that this is just above the normal temperature of the human body to which these organisms are adjusted. Beyond 450 there is a normal logarithmic temperature coefficient of about 1.2 per degree so that at 55 the average survival time has been reduced to 28 minutes. SUMMARY Experiments were conducted to determine the survival rates of bacteria which settled to the floor of a room after being sprayed into the air therein. The methods used were designed to simulate conditions obtaining when bacteria are sneezed or coughed into air by man. It was found that in a dark room the estimated time for 50 per cent survival for the three strains of streptococci studied (Group A and Group B beta-hemolytic, and alpha-hemolytic) varied from about one to thirteen days. In the case of one strain, 29B, the temperature coefficient of the death rate has the usual logarthmic form in the region between 45 and 55 degrees and value of about 1.2. Between 25 and 40 degrees the death rate is doubled which gives a coefficient of 1.047 if there be a coefficient of the usual logarithmic form. In the range 40 to 45 degrees there is an abrupt increase in the death rate amounting to about sixfold. 351 REFERENCE PHELPS, E. B., AND BUCHBINDER, L. 1941 Studies on microorganisms in simulated room environments. I. A study of the performance of the Wells air centrifuge and of the settling rates of bacteria through the air. J. Bact., 42, 321-344.