PUBLIC H E A L T H .
264
S P O T MAPS AND H O U S E I N F E C T I O N IN P H T H I S I S . ~ BY J. E . S A N D I L A N D S . M . D . . M e d i c a l Officer of H e a l t h for K e n s i n g t o n ,
I N voluntary tuberculosis
dispensaries organised on the Edinburgh lines one may see charts entitled " W h e r e Phthisis Breeds." These charts consist of squares representing houses. Each square contains a number of red dots, and the red dots indicate the number of phthisical patients who are known to have lived or to be living in the house. A square containing nine dots suggests, and I believe is intended to suggest, that nine, or at least eight, persons--if the primary case be deducted--in one house have become consumptive through the fact that they have come to live under this particular roof. The occurrence of these multiple cases in a house might be due to case-to-case infection, or to sanitary conditions in the house, such as lack of ventilation, which predispose to phthisis. The combined operation of both causes of phthisis may conveniently ba described by the inclusive and non-committal term, " H o u s e Influence." The m~in object of my paper is to suggest that these charts or spot maps, in the absence of a detailed analysis, are unscientific and misleading. During the last three years, that is to say, during a period extending from January ISt, I9tO, to November 24th , 1913, when the material for this paper was extracted from the registers, a record has been kept of the houses in Kensington in which persons notified as suffering from consumption are known to have lived. The data for the production of spot maps showing the incidence of phthisis on houses have thus been provided. For the purpose of analysis three essential additions have been made, namely a record in each case of the time during which the patient resided in the house, a record of the alleged date of the onset of the illness and a record in all multiple cases of those which appear to have been due to infection from contact with a consumptive member of the same family. The patients' house histories have not been taken back to any specific date. For each person notified, the history of previous changes of address, with dates, has not been noted beyond the period for which it was likely to be accurate. In most instances one previous address at least has been recorded, and three or four addresses *Read at a meeting of the Metropolitan Brauch of the Society of Medical Officers of Health on 27th February, ~914,
MAY,
have been noted where a patient has frequently changed his lodgings within a short period of the date of inquiry. The card register accordingly includes two classes of houses. The first class consists of houses where persons have resided whilst suffering from consumption ; the second class consists of houses where consumptives are known to have lived before the alleged date of the onset of their illness. On November 24th, 1913, the card register showed that there were in Kensington 1,895 houses where consumptive persons were known to have resided either before or during the course of t h e i r illness. Deducting eight common lodging.houses as not comparable'to the others, 1,887 houses remain. The incidence of single and multiple cases on these houses has been as follows : Single c a s e s o c c u r r e : l in 1,439 h o u s e s , or 76 p e r c e n t . M u l t i p l e ,,
.
448
Total
~,887
,,
,,
24
,.
ioo
,,
In the above table and in the context and tables which follow it must be understood that notified consumptives have been described and regarded for the purposes of the inquiry as " c a s e s " throughout their house history, which in many instances necessarily relates to their place of residence before they became consumptive. So far there have been no serious possibilities of fallacy, and we start with the solid fact that out of 1,887 houses in Kensington which have come under suspicion in connection with the notification of phthisis, 448 , or 24 per cent., have incurred the added stigma attending upon the occurrence of one or more secondary cases. In inquiring into the origin of these secondary cases, 27 liouses where the information obtained was insufficient have been eliminated. The houses remaining where the history of the multiple cases is known accordingly number 421 The first point to note is that in I92 of these houses all the secondary cases were due to family infection. In the best of houses and under the best conditions family infection may obviously occur. Further, such cases may be accounted for by heredity. On these two grounds the above 192 houses must be excluded from the number in which " H o u s e Influence" might be regarded as the chief factor in the production of multiple cases. Two illustrations may here be given of the kind of house influence to which the present inquiry is limited. One is the basement--so dark, damp, and ill. ventilated that its successive occupants almost
1914.
265
PUBLIC HEALTH.
invariably develop consumption; the other is the house where the consumptive lodger on the top floor spits about the passages and staircases and so infects the lodgers on the floors below. In another 86 houses the occurrence of more than one case proved, on inquiry, to have been due to chance alone. That is to say, either all the cases in each house, or all but the primary, were found to have been suffering from phthisis before their arrival. Finally, there were 41 houses in which chance and family infection each produced one or more of the secondary cases, and together accounted for the total number of cases in each house. In only lO2 houses did the dates of residence and onset of disease admit of the interpretation that house influence or house infection, as distinct from family infection, had resulted in the production of multiple cases. MULTIPLE
MODE
OF
Number
Per
of Houses.
cent.
Chance and Family Infection or Chance Alone . . . . . . . . . . . . . . . Chance and House Infection or House
127
55
lO2
45
229
Ioo
Infection Alone
. . . . . . . . .
Family Infection Alone
. . . . . . . . .
I92
Total ......
42I
Total ......
Less than I month ...... 1-2 m o n t h s . . . . . . . . . 2-6 months . . . . . . . . . 6 months
CASES.
Ot~IGIN,
is commonly produced by accident. And in these circumstances the mere fact that in lO2 houses the history admits of the possibility of house infection, does not to my mind afford any grounds for the assumption that house infection did in fact take place, or preclude the possible operation of chance in every instance. At the same time the evidence in favour of house infection and against chance should be considered. Any case in which the incubation period was less than t w o months must be regarded as weak. The following table shows, however, that the incubation period in the secondary case exceeded six months in 75 per cent., and wag only less than two months in 6 per cent. of the total cases : INCUBATION P E R I O D .
The above table shows that after excluding houses where multiple cases were due to family infection, 229 houses lemained in which a spot map would have indicated house infection. On further inquiry it has, however, been shown that in 55 per cent., or more than half these apparent cases of house infection, house infection did not in fact take place. Or in other words, the multiple cases in no less than 127 houses out of 229 were due to chance alone or a combination of chance and family infection. Neither do the figures 127 represent the total number of houses where one or more secondary cases have been due to nothing but coincidence. Among the houses where house infection may have occurred, there were 14 in which purely chance secondary cases were noted in addition to the possible infections. Coincidence accordingly operated in 141, or more than 6o per cent., of the 229 houses in which secondary cases not due to family infection occurred. There is accordingly abundant evidence that under the conditions of tenancy which prevail in certain parts of Kensington the phenomenon of the multiple invasion of houses by phthisis
or more
Nil. 6 cases, 2o ,,
...... Total
.,.
76
,,
io2
,,
It will be understood that the incubation period referred to is the period intervening between the onset of the disease in the second patient and his arrival in the infected house. Again, the primary case to be the origin of house infection should be moderately advanced and beyond the earliest stages of the disease. The following table shows the duration of the disease in the infecting case : DURATION O F D I S E A S E I N I N F E C T I N G C A S E . L e s s t h a n 6 m o n t h s in 6 months to 2 years in
... ...
years or more in ...... T o t a l d u r a t i o n u n k n o w n i n ...
ii cases. 28 ,, 56 7
,, ,,
IO2
In the above table the duration of the disease has been reckoned from the date of onset to the date on which the infected case developed the symptoms of phthisis. The duration cannot be reckoned to the actual date of infection, which is necessarily unknown. To this rule the following necessary exceptions have been made : Where the primary case had died or left the house before the onset of the disease in the secondary case, the duration has been reckoned up to the date of death or departure. Where the secondary case had left the infected house before developing the symptoms of phthisis, the duration of the disease in the first patient has been reckoned to the date of the departure of the secondary case. From the table it will be seen that in only i i per cent. of the total cases was the duration
PUBLIC H E A L T H .
66
of thedisease less than six months, and so short as to render it improbable that the primary case was not capable of transmitting the disease to others. In 84 per cent. of the cases the duration of the disease permits the first patient to be regarded as a potential source of infection. Apart from statistical considerations as to the possibility of chance association, the house histories, in the great majority of cases, show that house infection may well have actually occurred. In 25 instances the second patient was infected after the departure of the primary case from the same house or the same room. The intervening periods were as follows : ~ INTERVAL BETWEEN THE DEPARTURE PRIMARY CASE AND THE ARRIVAL
OF/file OF THE
LODGER IN THE SAME HOUSE O~ SAME
ROOM WHO PHTHISIS.
SUBSEQUENTLY
,H., NUMBER OF HOUSES.
INTERVAL ELAPSINC~.
Less than 2 months
Other Rooms Same House,
Same Room in House,
2
5
7
~
3
4
..,
~-5 m o n t h s . . . . . . 6-i2
montl~s . . . . . .
yezLr or m o r e Total
DEVELOPED
.., ..,
~.
Total.
2
o
2
3
9
12
8
17
25
~In five instances where the lodger developed phthisis after entering the room vacated by a consumptive patient less i:han two months beforehand, it is difficult to resist the conclusion that the source of infection did in fact reside in the room. W h e n the interval exceeded I2 months infection might be ruled out, but even so such cases could still be attributed to some baneful influence in the house or room, such as lack of light and air. In connection with this question of house influence as distinct from infection, it is pertinent to inquire as to t h e number of instances in which the primary case as well ~s the secondary case or cases fell ill after entering the house. The figures on this point: are as follows :-102 H O U S E S W H E R E HOUSE INFECTION APPEARS TO HAVE OCCURRED.
2 [
3 irotal~
H o u s e s w h e r e p r i m a r y case fell i ill before arrival ... .-. } 54 H o u s e s where p r i m a r y case fell I ill after arrival ...... ] 45
2
o
56
Totals . . . . . . . . .,., ,,,,
2
I
IO2
i~UMt3ER OF SNCON1)ARY CASES.
I
[
99
:
]
MAY,
It seems that if these Io2 houses deserved to be marked as breeding-grounds for phthisis, one would have at least expected the majority of their primary cases to have been home-bred, where:ts the figures show that 56 per cent. of the primary cases were imported. It remains to estimate the chances (vide summary in appendix) of any person who seeks a lo~gi~ng in Kensington happening to select a house Where a consumptive has lived or is living. There are 27,6oo inhabited houses in the borough, and 1,887 of this number are known within recent years to have harboured persohs who have been notified as suffering from consumption. The chance of a n y person selecting one of t h e s e " phthisis" housesis accordingly I in 15. Out of 1,887 phthisis houses rooms in 229, as w e have seen, have been selected as lodgings by persons now notified as suffering from consumption. The chance of a person, notified or to be notified as consumptive, finding a phthisis house is accordingly I in 8. His chance is twice as great as that of any casual seeker, and the inference is obvious. His connection with the phthisis house as a secondary notified case of consumption is not due to chance but to infection or "house influence." W i t h approximately equal populations in North and South Kensington, the deaths from phthisis are three times, and the notifications more than four times, as numerous in the northern half of the borough, where poverty prevails. In the North the number of phthisis houses is accordingly much greater, as is the number of such houseswhich are let in lodgings, and which are consequently available for the occurrence of multiple cases among different families under the same rooL Further, phthisis is a disease of the poor, and a person who is or is going to become phthisical will seek for lodgings in the North and not in the South. The chances of such a person finding a phthisis house should therefore be reckoned not on the total houses, but on the number of houses in the borough available for the classes who, being poor, are prone to consumption. I will not attempt to estimate the number, but I am able to produce proof that among dwellings of a certain class the proportion of phthisis houses is much greater than I in 15 and much greater even than I in 8. During the summer of 1913, diarrhoeawas made a voluntary notifiable disease in Kensington. Notifications were received from 4o2 houses and out of this number 119, or more than I in 4, were phthisis houses. At first sight it
1914.
PUBLIC HEALTH.
appears a reductio ad absurdum to have proved that diarrhoea is more frequently associated with phthisis houses than phthisis itself, but the explanation is not far to seek. Well-to-do persons were not notified, and the bulk of the diarrhoea notifications came from two doctors residing in N o r t h Kensington, and practising a m o n g s t the poorest classes in this part of the b o r o u g h . T h e diarrhoea register does not contain the address of a single goodclass house. T h e poor-class houses in the phthisis register, on the other hand, are leavened by a fair sprinkling o f genteel residences and commodious flats. Now all the latter class of residences should be excluded in reckoning the odds. T h e occupants retain their houses for 7, t4, or 2I years. Such houses are not let in tenements and are therefore not available, except at long intervals, for the house h u n t e r or, as has been said, for the occurrence of multiple cases. Personally, I do not feel a n y great strain in splitting the difference b e t w e e n I in 4 for the diarrhoea experience and I in 15 for all the houses in the borough, from the Palace downwards. T h e chances of a n y house hunter h a p p e n i n g on a phthisis house would then b e 2 i n l 9 o r I in 9½. T h e c h a n c e of the house h u n t e r who is or who is about to b e c o m e consumptive has been shown to be I in 8, and in view of the diarrhoea experience and the other points to which a t t e n t i o n has been called it seems to be not improbable t h a t this is no more t h a n t h e average chance of any house hunter in the class w h i c h provides t h e bulk of the entries in the notification and d e a t h registers for phthisis. On this assumption the whole case for house infection falls to the ground, and will not be made out until the registers show a v e r y m u c h heavier incidence of phthisis on phthisis houses. In conclusion, I must guard m y s e l f against misapprehension by a saving clause in place of a peroration. It has not been m y i n t e n t i o n to state or even hint t h a t house infection does not occur. W i t h those who go so far as to believe that i t occurs f r e q u e n t l y and is the cause of the majority of all cases of phthisis I have no quarrel. All t h a t I wish to point out is t h a t the figures and facts advanced, and s o m e t i m e s accepted, as p r o o f of house infection, do not in themselves prove a n y t h i n g of the k i n d . T h e y are worthless because t h e y are p r e p a r e d without regard to the possibilities of c h a n c e association, and in some instances more worthless still because they ignore the f u n d a m e n t a l point of the date of onset of the disease. And so houses
267
are put forward as the cause of phthisis in lodgers who have come there m o r i b u n d for a few m o n t h s on their way to the i n f i r m a r y . Finally, as against m y own figures and all others, if the infection in 90 per cent. of all cases is acquired in childhood, what bearing on the question of house infection has the fact that a lodger aged 35 came t o a phthisis house in I9Io, developed phthisis irk i912 , and died in I913 ? T h e disease has been assumed to have been acquired in the house in my inquiry if the onset of the illness was not less t h a n one m o n t h after the patient's arrival. But if the infection entered the patient's body 3o years before, of what value is a house history which c a n n o t extend over five years without going back beyond the date of the compulsory notification of phthisis into times which, in regard to the house incidence of phthisis, are prehistoric ? HOUSE INFECTION IN PHTHISIS. APPENDIX.
Houses in the B~rough. . . . . . . . . . tIouses where consumptives have lived Houses where multiple cases have occurred
..
27,6o0 1,887 229
Houses where house infection has occurred .. Houses where multiple ~ases were due to chance alone or chance and family infection . . . .
I27
TotaI houses where diarrheea was notified . . . . Phthisis houses where diarrhoea was notified
402 II9
io2
2~9
..
CHANCE OF LIVING IN A " P l l T I I I S I S H O U S E . "
A. For any person .... 1.887in 27,6o0 = i in 15. B. For a consnmptive . . . . 229 in 1,887 = x in 8. C. For a patient notified as suffering from diarrhoea ii9in 4o2 = I in 4. in I5 D. For a person who is poor { Ix in 4 e in 19 or I in 94
"SttUTTLIlg KlSSlNG."--Dealing with cases of pulmonary tuberculosis in Barnoldswick the medical officer of heaIth s a y s : - - " I attribute a large proportion of these cases to delicate people coming here to work from the neighbouring Lancashire towns. T h e y find the labour conditions here in the mill not exacting. The shuttle kissing in these weaving districts is, I have not the least doubt, the fertile source of communication, and I have tried to bring this home to the weaver, A circular letter was sent to each employer asking for co-operation, and the local councils passed a resolution supplying disinfectants to each mill. The recommendations are that whenever there is a change of weaver, shuttles should be thoroughly cleansed with the disinfectant, and all shuttles should be cleansed periodically,"