Clinical and epidemiological observations on an outbreak of plague in Nepal*

Bull. Org. mond. Sante| 1971,45, 693-706 Bull. Wld Hlth Org. Clinical and Epidemiological Observations on an Outbreak of Plague in Nepal* F. MARC LAFORCE,' I. L. ACHARYA,2 GORDON STOTT,3 PHILIP S. BRACHMAN,4 ARNOLD F. KAUFMAN,5 RICHARD F. CLAPP 6 & N. K. SHAH 7 In the autumn of 1967, plague broke out among hill people in western Nepal, a country that had not previously reported human plague. Two persons were infectedfrom an active sylvatic focus at a grazing area 5 km from Nawra, the village where the epidemic occurred. The second patient introduced plague into the village where the rest of the cases occurred. Clinical and epidemiological evidence suggests that plague was spread both by the airborne route, resulting in 6 cases of tonsillar plague and 1 case of primary pneumonic plague, as well as by infected fleas, resulting in 17 cases of bubonic plague. Since no evidence of a rodent epizootic was uncovered in the village itself, and because of the distinct clustering of the bubonic cases, human-to-human spread of plague by infected ectoparasite vectors, presumably Pulex irritans, is thought to have occurred. This focus probably represents the most southerly boundary of the central Asian plague area yet identified. Human plague was recorded as early as 1823 in India in the mountainous sub-Himalayan provinces of Garhwal and Kumaon (Renny, 1851), and this bacterial disease was apparently well established in these provinces up to 1900, since Simpson (1905) was able to collect reports of human plague in 25 of the almost 50 years. After 1900, plague became a major health problem in much of India, but the old foci in Garhwal and Kumaon became quiescent (Baltazard & Bahmanyar, 1960). The Kingdom of Nepal had never reported human plague until 1967, when cases were reported from Nawra, a small mountain village in the west- * From the Center for Disease Control, Public Health Service, US Department of Health, Education, and Welfare, Atlanta, Ga., USA; the Ministry of Health, Government of Nepal; and the World Health Organization. 1 EIS officer, Epidemiology Program, Center for Disease Control. Present address: Fellow in Infectious Diseases, Channing and Thorndike Memorial Laboratories, Boston City Hospital, Boston, Mass., USA. 2 Medical officer in Infectious Diseases, Bir Hospital, Kathmandu, Nepal. 3WHO Representative for Thailand, c/o Ministry of Public Health, Devavesm Palace, Bangkok, Thailand. 'Director, Epidemiology Program, Center for Disease Control. 6 Veterinary Pathologist, Bacterial Diseases Branch, Epidemiology Program, Center for Disease Control. 6 Chief, Sanitarian Development Activity, Health Profes- sions Branch, Training Program, Center for Disease Control. I Assistant Director, Department of Health, Singha Durbar, Kathmandu, Nepal. ern part of the country. Because of the possibility of spread, an epidemiological investigation was undertaken to define both the animal reservoir and the extent of the human outbreak, and to institute appropriate control measures to arrest the spread of the disease. This paper contains pertinent clini- cal and epidemiological data obtained during this investigation. The newly discovered focus was found to be related to plague in the old Indian provinces of Garhwal, Kumaon, and Uttar Pradesh. On 30 October 1967, the Ministry of Health of the Government of Nepal was notified of an out- break of illness with fatal cases in man and cattle in the village of Nawra, Bajhang District. The fol- lowing day, representatives of the Ministry of Health, the World Health Organization, and the US Agency for International Development (USAID) flew to Nawra to make a preliminary investigation. Bacteriological specimens were obtained and pa- tients ill at the time were examined. An initial diag- nosis of anthrax was based on the cutaneous lesions seen, the reports of large numbers of cattle dying, and the absence of rodent deaths. Patients were given penicillin therapy and the epidemiological investigation was started. On 8 November, the investigators were joined by a group from the Na- tional Communicable Disease Center, now Center for Disease Control (CDC), Atlanta, Ga., USA. It soon became apparent that the cattle deaths 2761 - 693 F. M. LAFORCE AND OTHERS were due to rinderpest and that plague was the cause of the human illnesses and deaths. THE VILLAGE Nawra, a Hindu village, is located on the eastern slope of a 2900-m mountain in far western Nepal about 64 km east of India and 64 km south of Tibet (Fig. 1). It is relatively inaccessible, being 5-7 days by foot from the nearest light aircraft landing strip at Doti or 30 minutes by helicopter. The village, situated at an altitude of 1980 m, forms part of a group of villages (Subeda Pan- chayat) scattered over an area 5 x 8 km on a ter- raced mountainside extending from the Seti River, altitude 914 m, to the village of Surke Mela at 2590 m (Fig. l). Chainpur, the district head- quarters, is to the north-east of Nawra, a 1-2 hours' walk away. The area was affected by earthquakes and drought 2 years previously, and the agricultural economy was at only subsistence level. Subeda Panchayat has a total population of 1500 people, and Nawra has about 650 inhabitants. The people are farmers and live in two-storey houses made of stone covered with a thin layer of mud and lime, the roofs being stone or straw. The first floor serves as a cooking area and as a living-room, the second floor is used for sleeping; cattle, buffalo, or goats are housed on the ground floor. Rooms are generally windowless and small, and most houses are partitioned to accommodate several families or households. MATERIALS AND METHODS Swabs from skin lesions and bubo aspirates from persons suspected of having plague were plated on blood agar, and swabs were placed in nutrient broth. Plating was carried out in the field whenever possible. The plates were examined after 48 hours and suspicious colonies were stained by Gram's method, subcultured, and then grown in nutrient broth; equal samples were then injected into mice. Pure bacterial cultures were obtained from the mice that died, and these cultures were identified as Pasteurella (Yersinia) pestis by stan- ..SURKE MELA -. (8 500 feet; 2 590 m) I THALABASTI DHANAU RANUTBADHA r INDIA TIBETKAR 0NDOTIBET KATI LUM INDIA- 0 50 100 150 miles : !0 200 km D L WHO 15643 Fig. 1. Subeda Panchayat area, Nepal, showing the two active plague sites of Surke Mela and Nawra. SALLYANE -TRAIL BHOGRE GWANE DEWAL 694 OBSERVATIONS ON AN OUTBREAK OF PLAGUE IN NEPAL dard cultural, biochemical, and serological methods During the course of the investigation, 10 fleas (Baltazard et al., 1956). were removed from dogs at Nawra; 1 flea was Human and canine blood samples were centri- identified as Pulex irritans and the other 9 fuged within 12-24 hours of being taken, and the belonged to the species Ctenocephalides felis. The serum was frozen. The serum specimens were then 10 fleas were pooled and inoculated into mice; tested by the microtitre modification of the passive no recovery of P. (Y.) pestis was made. haemagglutination test for P. (Y.) pestis fraction 1 specific antibodies (Chen & Meyer, 1966). CLINICAL DETAILS Tissue impressions were taken, and a smear of one bubo aspirate was made; these were all exam- On the basis of epidemiological, microbiological, ined by a fluorescent antibody staining technique and clinical evidence, it was recognized that 26 cases (Moody & Winter, 1959). Tissue specimens were of human plague occurred in the village of Nawra also inoculated into mice. between 6 September and 5 November 1967. Table 1 Table 1. Clinical characteristics of plague cases Case Name No. of patient 1 Laxmi 2 Sanke 3 Dharam S. 4 Bal B. 5 Lalsara 6 Rupa 7 Gogan 8 Dhabbal 9 Thule S. 10 H arka 11 Ramchandra 12 Jaldhara 1 3 M ata 14 Tulsi 15 I Kalu S. 16 Man Bir 17 Laxmi 18 Nurma 19 Nani Koila 20 Janakala 21 Jhupri 22 23 24 25 26 Keshu Hastia Shiva K. Ramgiri Lware S. 1 AgeS x 1 House Family] of Date of 1Selns1Skin PulmonaryAge Sm iY onset (1967) death (1°967) Swellings lesion symptoms a I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 16 24 22 30 40 12 6 15 40 15 5 30 15 50 50 6 8 60 13 20 13 7 25 11 62 F M M M F M M M M M F F F M M F F F F F F M F M M A B G B B B B C B B B B E E D B E E D B c B B C B B 1 2 11 3 2 2 3 6 2 4 3 3 9 9 7 4 9 9 7 2 4 4 4 6 3 2 6 sept. 24 Sept. 28 Sept. 2 Oct. 5 Oct. 5 Oct. 5 Oct. 5 Oct. 6 Oct. 10 Oct. 11 Oct. 12 Oct. 14 Oct. 18 Oct. 19 Oct. 20 Oct. 22 Oct. 23 Oct. 24 Oct. 25 Oct. 25 Oct. 25 Oct. 26 Oct. 26 Oct. 28 Oct. 5 Nov. 8 sept. 28 Sept. 7 Oct. 7 Oct. 9 Oct. 14 Oct. 12 Oct. 9 Oct. 12 Oct. 15 Oct. 15 Oct. 18 Oct. 20 Oct. 28 Oct. 28 Oct. 28 Oct. 28 Oct. 3 Nov. axillary inguinal neck, axilla cheek, trunk neck trunk neck axilla trunk neck shoulder cheek axillary supraclavicular inguinal neck trunk supraclavicular, inguinal b neck axillary neck inguinal foot hand leg head trunk shoulder b calf b neck neck b leg c B-BI B-BI B-BI B-BI B B-BI B-BI BI C C B-BI B C c B-BI a B = black sputum; BI - bloody sputum; C = cough. b p. (Y.) pestis recovered. 695 F. M. LAFORCE AND OTHERS and Fig. 2 summarize the pertinent clinical and demographic data for each case. Among the 26 cases there was an equal sex distribution and 69% of the cases occurred in the 1-29-year age group. Altogether, 17 of the 26 patients died before the investigation began. The course of illness in these 17 persons was established by repeated interviews with family members, friends, or other responsive villagers. The clinical picture was strikingly similar in almost all cases, consisting of a sudden onset of fever and chills and the appearance of painful swell- ings in the facial, cervical, axillary, inguinal, or truncal areas. The illness continued with high fever, and in 14 of the 17 patients the appearance of pul- monary symptoms, usually characterized by the expectoration of black and bloody sputum. Vomit- ing was noted in 12 patients and diarrhoea in 8. Death occurred 3-10 days after the onset of symp- toms. Two of the patients were reported to have had multiple " black spots" scattered over the skin prior to death. The significance of the clinical picture in this group is discussed in greater detail later. The 9 patients who were examined by members of the investigating team had all had abrupt onset of fever with regional lymphadenopathy; 8 of CLINICAL DATA CASE NUMBER LESION I SWELLING SME MELA. FOOT HAND LEG HEAD TRUNK SHOULDER & CALF NECK NECK LEG 2 AXILLARY 3 INGUINAL 4 AXILLARY NECK 5 TRUNK a CHEEK 6 NECK 7 TRUNK 8 NECK 9 AXILLARY 10 TRUNK II NECK 12 SHOULDER 1S CHEEK 14 AXILLARY 15 SUPRACLAVICULAR 16 INGUINAL 17 NECK is 19 TRUNK 20 SUPRACLAVIULAR 21 NECK 22 AXILLARY 23 24 NECK 25 INeUINAL 26 the 9 had skin lesions that ranged from small, hard, dark eschars to large ulcerations. Three representa- tive cases are described in detail. Case 21 Jhupri, a 20-year-old woman from house B, first had fever on 25 October. Small painful lesions developed 3 days later on the left shoulder and right calf; they enlarged, ulcerated, and developed black central eschars. When examined on 31 Octo- ber, the patient was febrile (102.6°F; 39.2°C) and toxic. A 3-cm ulcer with a black eschar and red or edematous margins, originally suggestive of cuta- neous anthrax, was seen on the left shoulder. A 4-cm shallow ulcer on the right lower calf was also noted, as well as tender right inguinal lympha- denopathy. The liver was palpable 2 finger-breadths below the right costal margin, and there was dull- ness to percussion over the left lower lobe, with rales in the same area. The right lung was clear. Therapy with parenteral penicillin (800 000 units per day) was initiated on 31 October but the patient did not improve. On 2 November her temperature was 102°F (38.9°C) and coarse rales were still pres- ent in the left lower lobe; tetracycline (1 g per day) was included in the treatment. On 4 November the patient was afebrile but a large left supraclavi- EPIDEMIOLOGICAL INYESTIGATION NEPALESE AND WHO TEAM HOUSES SPRAYED WITH PHENOL SOLUTION II|CDC TEAM IiI----+ __ + ------I I - - -~~~~~~-lI_ __~~~~~~- 4-r 10 24 2S 2 2 4 IZ 14 1 11 22 24 26 2630 I 5 7 SETEMSER OCTOBER NOVNER Fig. 2. Flow of plague through Surke Mela and Nawra from 6 September to 11 November 1967. Asterisks (*) indicate positive serology for, or isolation of, Pasteurella (Yersinia) pestis. (The squares indicate the date of onset of illness and the letters within them identify the house involved; shading indicates cases with pulmonary symptoms; dashed lines indicate the duration of illness of fatal cases, each dash representing one day and the cross the date of death; solid lines indicate non-fatal cases.) 696 t_ --. N^wNZ -tTI ?4 a1 0 I= M------ - - - - - - - M__ OBSERVATIONS ON AN OUTBREAK OF PLAGUE IN NEPAL cular bubo had developed. During the next 5 days she showed some improvement but the bubo re- mained large and tender and became fluctuant. The patient started taking streptomycin on 12 November (1 g per day) and subsequently showed improvement. Swab cultures from the shoulder and calf lesions, as well as aspirate from the right inguinal node, were all positive for P. (Y.) pestis. Comment. This case illustrates the clinical syn- drome of primary plague carbuncle, an unusual skin manifestation of plague that was originally mistaken for cutaneous anthrax. Plague carbuncles were first described by Chosky (1909) as " cellulo- cutaneous plague" and were found in 3 % of 13 600 cases of bubonic plague seen at the turn of the century in Bombay. This lesion was noted in 3 of the 9 patients examined by members of the team (cases 17, 19, 21), but all three recovered. The development of plague carbuncles has previously been reported to be associated with a favourable outcome of the disease (Machiavello, 1951). Case 22 A 13-year-old girl from house B, Keshu, became ill on 26 October with fever and swelling in the right lower part of her neck. When examined on 31 October she was afebrile and had a pea-sized pustular skin lesion in the right suboccipital area, with small tender posterior cervical nodes. She was given procaine penicillin (400 000 units intramus- cularly twice a day), but her condition became pro- gressively worse over the next 3 days with oral temperatures of 102-104°F (38.9-40'C). Oral tetra- cycline (1 g daily) was added to the treatment on 2 November and continued until 9 November. On 7 November pronounced nuchal rigidity devel- oped and the patient lapsed into semicoma during the next day. Chloramphenicol (1.5 g intramus- cularly daily) was substituted for the tetracycline on 9 November and continued until 15 November, at which time the patient began a 10-day course of parenteral streptomycin. Parenteral fluids were administered daily to maintain hydration, and over the next 2 weeks the patient slowly improved and completely recovered. A serum specimen drawn on 31 October was positive (1: 128) for antibodies against fraction 1 of P. (Y.) pestis. Comment. This case was undoubtedly one of secondary meningeal plague. The equipment neces- sary to perform a lumbar puncture was not avail- able but the clinical picture and the positive serology confirmed the diagnosis. The development of men- ingeal signs 12 days after the onset of illness is consistent with the pattern of secondary meningeal plague in adults as described by Landsborough & Tunnel (1947). In their series of 8 cases, central nervous system signs appeared 9-17 days after the onset of illness. Although considered unusual, Martin et al. (1967) reported the meningeal spread of plague in 3 Navajo Indian children during the 1965 outbreak of 7 cases in south-western USA. Case 26 Lware S., a 62-year-old man from house B, first noticed a small skin lesion on the inner surface of his right thigh on 5 November; 3 days later he was febrile (102.6°F; 39.2°C) and had right inguinal adenopathy. The patient was started on procaine penicillin (400 000 units intramuscularly twice a day); he did not improve, and on 12 November after a shaking chill he was noted to be disoriented and ataxic. Examination at that time revealed a small, leathery eschar on the medial surface of his right thigh with inguinal bubos both above and below Poupart's ligament. An attempt to aspirate the bubos was unsuccessful, and he was started on streptomycin (1 g intramuscularly daily); within 2 days he was afebrile and apparently well. Serum drawn on 14 November was positive (1: 640) for antibodies against fraction 1 of P. (Y.) pestis. Comment. This case is one of classic bubonic plague and illustrates the second type of skin lesion, which was seen in 5 of the 9 cases examined. The lesion, characteristically pruritic early in its course, formed a small vesicle, which ruptured, leaving in its place a small, leathery eschar that slowly healed. In all cases where this lesion was seen, peripheral adenopathy with bubo formation also occurred. LABORATORY RESULTS The laboratory results in cases of human plague at Nawra are summarized in Table 2. In 4 cases, appropriate bacteriological material was taken; 4 isolates of P. (Y.) pestis were obtained from 2 patients. Initial bacteriological investigations and identifications were conducted by a WHO laboratory technician assigned to the Department of Health, Public Health Laboratory, Kathmandu. Confirmatory laboratory studies were carried out at the All India Institute of Medical Sciences, New Delhi, the Haffkine Institute, Bombay, and the 697 698 F. M. LAFORCE AND OTHERS Table 2. Laboratory results from plague cases Serology Case No. Bacteriology S l Date (1967) Titre 3 no sample 14 Nov. 1:32 15 negative culture from bubo aspirate 31 Oct. negative 14 Nov. contaminated 19 no sample 14 Nov. negative 21 positive cultures from skin lesions and bubo 31 Oct. negative 14 Nov. 1:4 096 22 negative culture from neck lesion 31 Oct. 1:128 25 positive culture from neck lesion as well as from no sample post-mortem specimens of sternal marrow and a digit 26 no sample (bubo aspirate unsuccessful) 14 Nov. 1 64 National Institute of Communicable Diseases, New Delhi, India. As previously indicated, plague bacilli were isolated from the shoulder and ankle lesions as well as from an inguinal bubo in case 21. P. (Y.) pestis was also recovered from a neck lesion in case 25. Additionally, post-mortem spe- cimens of sternal marrow as well as a digit from case 25 were positive for P. (Y.) pestis. The isolates from both these patients fermented glycerol and reduced nitrates. Serum specimens were positive in 4 of the 6 cases in which the serum obtained could be tested. One of the positive specimens was from a villager (case 3) who was originally thought not to have plague but was bled as part of a group of 8 persons who had varying degrees of contact with affected persons. The titres of the other 7 members of this group were negative. A total of 23 other villagers of Nawra, chosen at random, were bled on 14 No- vember 1967. All these serum specimens were tested and found negative. Two dogs from Nawra and 2 dogs from Surke Mela were bled, and signifi- cant titres to fraction 1 of P. (Y.) pestis were demonstrated in 3 of them (Table 3).1 Specimens from sternal marrow from case 25, as well as purulent material aspirated from a fluc- tuant bubo from case 21, were also positive when 1 Since this article was submitted for publication, Rust et al. (1971a, 1971b) have studied the serological responses of dogs and cats to P. (Y.) pestis and have also demonstrated the usefulness of screening canine sera for antibodies to P. ( Y.) pestis as a marker of plague activity. Table 3. Serological studies of domestic dogs in the plague epidemic area Location No. of dogs I HA Titre a Surke Mela 1 1:16 1 1:128 Nawra 1 Negative 1 1:32 a Haemagglutination titre for fraction 1 specific antibodies. tested with anti-fraction 1 fluorescent antibody con- jugate. On 12 and 13 November the clothes of all per- sons living in affected houses were examined for fleas, and in addition all the affected homes were swabbed for fleas with large cotton strips. No fleas were recovered either from the clothing or the homes, perhaps because phenol solution had been sprayed liberally during the previous week. EPIDEMIOLOGICAL INVESTIGATION Although all human cases of plague in this out- break were in Nawra villagers, the first two cases appear to have been related to village activity at Surke Mela, a grazing area about 5 km south-west of Nawra at an elevation of 2590 m. A map of the epidemic area in Nawra with OBSERVATIONS ON AN OUTBREAK OF PLAGUE IN NEPAL ID EO C3. O IZELE] LIE] l 2 -. ~~~78 LIE]~~~~~~~~~~~~ SURKE MELA E v\ ! THEOKO GRTAVEEYARD .- 6 55 ., A 02"," .! 8 4 3 2 H: ..12,,' Fig. 3. Epidemic area in village of Nawra; letters refer to the house and numbers to families. (Open squares: un- affected households; solid squares: households with cases of plague; shaded squares: households with contacts but with no infections; dashed lines: paths.) house and family identifications is shown in Fig. 3. The first case of illness occurred in Laxmi, a 16-year- old girl from house A, who was at Surke Mela tending cattle when she became ill on 6 September. Her illness was characterized by fever, cough, diar- rhoea, vomiting, abdominal pain, and haema- temesis. No swellings or skin lesions were reported. She died unattended at Surke Mela on 8 September and was wrapped in shrouds, taken down the moun- tain, and cast into the Seti River. Of the 7 persons who came in contact with her during her illness and burial ceremonies, all remained well except for a cousin, Dharam S. (case 3), who became febrile and had axillary swelling but recovered without treatment. Initially he was thought not to have plague because of the mild nature of his illness, but a serum specimen drawn on 14 November was positive (1: 32) for antibodies against fraction 1 of P. (Y.) pestis. Two weeks after Laxmi's death, Sanke, a 24-year- old man living in house B, went from Nawra to Surke Mela to look after the cattle. He arrived on 21 September and stayed in the house where Laxmi had been ill and had died. He became ill on 24 September and 2 days later returned to house B in Nawra. He was said to have fever, cough with bloodstained sputum, diarrhoea, and vomiting. No swellings or skin lesions were re- ported. He died in Nawra on 28 September, the fifth day of his illness. House B, where the second death occurred, was shared by 4 families, each with separate accom- modation. One of the four families, family 5, com- posed of 2 individuals, left the main house on 7 October to work on the rice crop. They returned on 18 October but went to house C, presumably because of the serious illness then sweeping through house B. Thus, there were 3 families comprising 15 members who lived in house B throughout the course of the epidemic. Seven days after the retlirn of Sanke to house B, Bal Bahadur from family 3 developed fever, cough, blackish bloodstained sputum, and swelling on the right foot and right inguinal region. On 5 October, 3 more cases developed in house B and one in house C. On 6 October there was yet another case in house B. The 5 cases that occurred on 5 and 6 October were all clinically similar; all patients LI 699 F. M. LAFORCE AND OTHERS had cough with black and bloody sputum, all had swellings, and in 4 of the 5 cases the swellings were in the facial or cervical area. New cases continued to occur almost daily in house B, and subsequently the disease spread to the adjoining houses D and E. House D, close to house B, was shared by 2 separate households; one had 2 cases of rather mild illness while the other remained unaffected. A family of 8 lived in house E; 4 became ill, and 2, both of them children, died. On 18 October, after the second case occurred, the family in house E moved to a house higher up the mountainside 4. TOTAL CASES 3. 21 7HI (Fig. 3, E 1). One of the two persons in family 5, who left house B and returned to house C on 18 October, became ill on 25 October. Over an 8-week period, 9 households were affected in 5 separate houses; 26 of the 40 members became ill, and 18 died (Fig. 4). The epidemic was most severe in house B. There, the disease appeared to occur in two separate waves, the first on 5 and 6 October and the second from 20 to 28 October. Cases in the surrounding houses occurred between 28 September and 26 October. Attack rates and case fatality rates for individual houses are given in Table 4. The attack rate (100 %) Fl 0 w a N2 UO c w z D -0 It J HOUSE A ]J HOUSE C 0oL ~ 1I IIh1 LW 1P :I HOUSE e HOUSE a 6 8 10 24 &B41&0Z''w6 8 1 Z 4 1 0 '2 24 '6 '6 30 1 & t SEPTEMBE OCTOBER NOVEMBER Fig. 4. Epidemic trend of plague in Nawra from 6 September to 5 November 1967, by date of onset. (Each open rectangle indicates a non-fatal case, each shaded rectangle a fatal case; S = swellings; P = pulmonary symptoms.) M H E L - in-irFT1 700 OBSERVATIONS ON AN OUTBREAK OF PLAGUE IN NEPAL Table 4. Plague attack rates and case fatality rates by household No. of Total Cases of plague Deaths from plagueHouse No.seofd TotalrHousehouseholds members No. Percentage No. Percentage A 1 6 1 17 1 100 B 3 15 15 100 13 86 c 2 5 3 60 2 67 D 1 4 2 50 0 0 E 1 8 4 50 2 50 G 1 2 1 50 0 0 total 9 40 26 65 18 ] 45 and the case fatality rate (86%) in house B were significantly higher when compared with the totals for houses A, C, D, E, and G. In addition to household contacts of cases, another 9 persons had varying degrees of direct contact. One of these contracted plague (case 3) and has already been described. There were no cases in a second group of 34 probable contacts, that is, persons who may have come in contact with an affected person or visited an affected home. Thus, the most critical factor determining the likeli- hood of infection was inhabiting, or contact with, an affected house (specifically, house B). When the cases were analysed for the presence or absence of pulmonary symptoms and outcome, an impressive relationship between the presence of pulmonary symptoms and mortality was seen (Table 5). In all, 15 of the 18 patients who died had pulmonary symptoms, while only 1 of the 8 survivors had pulmonary symptoms. When the patients with pulmonary symptoms were further divided by house, 12 of the 15 resided in house B. Thus the higher proportion of cases with pulmonary involvement accounted for the greater case fatality ratio seen in house B. During the course of the investigation, 6 other villages were visited. There were no plague-like illnesses among these people. Villagers could not Table 5. Pulmonary symptoms in fatal and non-fatal cases Fatal cases Non-fatal cases Hue Pulmonary puNo Total Pulmonar No Total Isymptoms pulmmptonary deaths symparyt pulmonary survivorssymptoms symptoms A 1 0 1 0 0 0 B 12 1 13 0 2 2 c 2 0 2 1 0 1 D 0 0 0 0 2 2 E 0 2 2 0 2 2 G 0 0 0 0 1 1 total 15a 3 18 1 b 7 8 a 11 with "black and bloody " sputum, 4 cough alone. b Cough only. 701 F. M. LAFORCE AND OTHERS recall any previous outbreaks or single cases that resembled the illness seen at Nawra. Likewise, no cases of similar disease were reported to the district headquarters at Chainpur. Persistent and detailed inquiries at Nawra failed to elicit any information suggestive of an epizootic in " rats " or other " rodents ". However, the father of the first patient said that he had seen 3 dead " rats " at Surke Mela in August. This was corro- borated by a person who lived next to the stable where the first two patients slept, who stated that he had seen a dead " rat " at the stable after Laxmi's death. No rodent deaths were observed in Nawra or the 6 other villages visited during the course of the investigation. CONTROL MEASURES From the onset of the investigation the quaran- tine measures, originally formulated by the villagers themselves, were intensified and continued. Addi- tionally, as a general control measure, the floors, walls, and ceilings of all the houses in Nawra were sprayed daily on 7-9 November with a 5% solu- tion of phenol. The houses where illness had occurred were sprayed 3 or 4 times. Subsequent to confirmation of the diagnosis in the period 13-18 November, 640 villagers in Nawra and 197 people living in a nearby village were started on prophylactic sulfaphenazole (Orisul; 0.5- 1.0 g orally) or streptomycin (1 g intramuscularly daily) for 7 days. In addition, all houses in Nawra were resprayed with residual DDT (1-2 g/m2) by a team from the Nepal Ministry of Health malaria eradication pro- gramme. DISCUSSION Mode of spread ofplague in Nawra A review of the clinical characteristics and tem- poral course of plague through Surke Mela and Nawra (Fig. 2) suggests and supports two separate, but related, epidemiological events with regard to the mode of spread and the clinical forms of the disease. The first wave, which was limited mainly to house B on 5 and 6 October, was characterized by 4 fatal cases, all with histories of pulmonary symptoms that included black and bloody sputum. These findings are suggestive of primary pneumonic plague where the index case could have been case 2 who became ill at Surke Mela and died with pulmo- nary symptoms 2 days after returning to Nawra. However, against the hypothesis of primary pneu- monic plague during this first wave are the histories of " swellings " in various places such as the neck and cheek, a sign not associated with primary pneu- monic plague (Wu Lien-teh, 1926). Furthermore, the average time between the onset of symptoms and death was 51/2 days, which is longer than the duration usually associated with primary pneu- monic plague (Chun, 1936). A second possibility that could account for this first wave of cases is the occurrence of bubonic plague with the development of secondary plague pneumonia. Against the diagnosis of classic bu- bonic plague in the early cases in house B is the high frequency of swellings in the face and neck and essentially no swellings associated with the extremities. However, the first cluster of cases in house B is consistent with the syndrome of tonsillar plague. This type of plague usually occurs as a result of aerosol transmission of large droplets containing P. (Y.) pestis to the tonsils or pharynx, with the development of cervical bubos and secondary pneu- monia. Experiments in primates by Meyer & Larson (1959) indicate that the site of deposition of P. (Y.) pestis in the respiratory tract determines the clinical syndrome. Large particles containing plague bacilli are filtered in the nasopharynx and lead to the syndrome of cervical or tonsillar plague, while particles smaller than 5 ,um are deposited in the lower respiratory tract and cause primary pneumonic plague. Although airborne spread is postulated as the mode of transmission in the early cases in house B, all but one later case (case 24) at Nawra were classically bubonic and therefore probably involved a flea vector. The single exception (case 24), a 25-year-old woman who lived in house C, suddenly became ill on 26 October, developed black and bloody sputum, and died on the third day of her illness. She had no reported skin lesions or swellings. Her clinical history is entirely compatible with a diagnosis of primary plague pneumonia. The source of infec- tion, while unknown, could have been case 16, a 50-year-old man from house B, whose illness was characterized by the expectoration of black and bloody sputum. No further cases of primary pneumonic plague resulted from this case, probably because of enforced quarantine. Case 24 was cared for by a 60-year-old member of house C who was never ill although he associated intimately with members of both houses B and C throughout the entire course of the epidemic. 702 OBSERVATIONS ON AN OUTBREAK OF PLAGUE IN NEPAL Since the classic work of the Indian Plague Com- missions, the rat flea, Xenopsylla cheopis, has been thought to be the only important vector in the transmission of plague. However, Blanc & Baltazard (1945), after noting the familial, domiciliary nature of plague outbreaks in Morocco, carried out exten- sive laboratory experiments and proposed that human ectoparasites were of great importance in the epidemic spread of plague. This concept was further broadened and documented in an extensive review by Blanc (1956). While this concept has not been accepted by some authorities (Girard, 1943), others have agreed that interhuman transmission of plague by infected human ectoparasites may play an important role in certain areas (Swellen- grebel, 1953; Pollitzer, 1954). Several factors in this outbreak point to the like- lihood of transmission of plague by infected human ectoparasites from man rather than spread by ecto- parasites from rodents. First, the focal nature of the outbreak was striking. Second, all cases of plague could be traced to a contact with an earlier case of recent date. Finally, no evidence of a rodent epizootic was found in Nawra despite persistent inquiry and careful searching. The flow of plague through house B may be con- strued retrospectively as having possibly proceeded through an initial airborne phase followed by con- tinued interhuman transmission by fleas (Fig. 5). The first patient from house B contracted plague at Surke Mela presumably from " rodent " ecto- parasites. The presence of two " rodent " carcasses in late August and September, as well as the posi- tive dog serology, testify to the presence of plague at Surke Mela. On the patient's return to house B at Nawra, he developed secondary plague pneu- monia and 4 persons who had contact with him contracted what is presumed to have been tonsillar plague. Between 7 and 15 October there were 8 deaths in house B. It is possible that, as a result of ter- minal septicaemias, the number of infected human Fig. 5. Spread of plague in house B, Nawra. (Numbers refer to the case number; crosses: deaths; solid squares: tonsillar plague; open squares: bubonic plague; horizontal lines terminating in short vertical lines: estimated time of inoculation.) 2 2 *N ** t 4 :P5 h- 1 t 7 v@t 10 Ee .J*t HOUSE SPRAYED WITH PHENOL SOLUTION EXTRINSIC INFECTED FLEA POOL In INCUATION 16 20 1__q 2-@ 23 1 1A 2a 25 El . t 26 ....El.... l I _ 24 26 28 30 2 4. 6 8 10 12 14 16 18 20 22 24 26 28 30 3 5 7 9 11 13 SEPTEMBER OCTOBER NOVEMBER I1967 H. 10647 703 F. M. LAFORCE AND OTHERS fleas increased in house B. Subsequent cases in house B, all of the bubonic variety, probably reflected continued interhuman transmission via human fleas. The cases in other houses, which include 1 case of tonsillar plague (case 14), 1 case of primary pneumonic plague (case 24), and 8 cases of bubonic plague, represent contact with persons or fleas from house B. This infected flea pool may have been significantly diminished as a result of vigorous spraying with phenol. This hypothesis of the mode of spread may help to explain why the outbreak remained contained. First, quarantine arrangements limited human- human contact and stopped the airborne spread of plague. Second, the quarantine of affected houses prevented susceptible persons from entering areas, such as house B, where some of the fleas may have been infected. Although the climate and living conditions in Nawra would favour high densities of human ecto- parasites (Swellengrebel, 1953) no studies of ecto- parasites were made. Therefore, the role of inter- human transmission of plague by human ectopara- sites in this outbreak, while consistent with the epidemiological findings, must remain in the final analysis speculative. Serological results in dogs The finding of significant antibody titres to frac- tion 1 of P. (Y.) pestis in 3 of the 4 dogs sampled is impressive. In the 1965 outbreak of plague among Navajo Indians in the USA, domestic dogs in the vicinity of the 3 human cases were found to have positive serology (Kartman et al., 1967). The dogs at Surke Mela and Nawra were not fed by villagers but foraged for themselves. Of necessity this placed them in intimate contact with the local animal population where they were prob- ably exposed to P. (Y.) pestis. Thus, dogs may serve as excellent sentinels of plague activity among the resident animal population. Unfortunately, lit- tle work has been done on experimental infection and antibody response to plague in canines, but it is not inconceivable that large-scale sampling of domestic dog sera, being convenient to carry out, might become a valuable epidemiological technique in mapping areas of plague activity. Relationship of the Nepal outbreak to plague in India The two strains of P. (Y.) pestis isolated from patients in Nepal fermented glycerol and reduced nitrates. Since all strains of P. (Y.) pestis isolated in India have been glycerol-fermentation negative, the Nepal strains must be considered to be unre- lated to plague foci in Uttar Pradesh, the Indian Province contiguous with Nepal. Rather, they sim- ply represent what seems to be the most southerly tier of the ancient Central Asian seat of plague yet described (Raettig, 1961). This finding offers concrete evidence in support of the hypothesis proposed by Baltazard & Bahmanyar (1960), after extensive field investigations in Uttar Pradesh, that the unstable plague foci in that province were not related to the ancient Indian sub-Himalayan foci of Garhwal and Kumaon. Century-old descriptions by Renny (1851), Francis & Pearson (1854), and Francis (1879-80) of out- breaks of plague in the mountains of Garhwal and Kumaon are of more than historical interest. First, the descriptions of the mountain villages and their inhabitants could easily be substituted verbatim for what was seen by the present authors in Nepal. Second, Francis (1879-80) made specific mention of great interhuman communicability of bubonic plague, as was the case in the Nepal outbreak. Thus, it is probably true that western Nepal represents simply part of the southernmost boundary of the ancient plague focus of Central Asia, to which the old Indian provinces of Garhwal and Kumaon belong. It is likely that sporadic cases of plague have occurred in western Nepal in the past. The isola- tion of villages because of the mountainous terrain would virtually guarantee that isolated outbreaks of plague would go unnoticed among the usual " fevers " that afflict these hill people. On the other hand, it is probably safe to assume that, at least in Bajhang District, plague has never struck with the severity of the 1967 epidemic. ACKNOWLEDGEMENTS The authors gratefully acknowledge the assistance of the following persons without whose generous collabo- ration this investigation could not have been completed: Dr Y. R. Joshi, Director, Department of Health, Min- istry of Health, Government of Nepal; Dr N. R. Joshi, Veterinary Officer, Central Veterinary Hospital, and 704 OBSERVATIONS ON AN OUTBREAK OF PLAGUE IN NEPAL 705 Dr D. D. Joshi, Veterinary Officer, Doti, Department of Veterinary Medicine, Government of Nepal; Mr J. B. Dhaubhadel, Senior Sanitarian, Malaria Eradication Branch, Government of Nepal; Miss J. R. Nelson, Laboratory Technician, Mr C. Seetharaman, Animal Health Officer, and Dr B. B. G. Nehanl, Microbiologist, World Health Organization; Mr J. S. Benz, Director, Mr J. C. Flint, Deputy Director, Dr L. Florio, Regional Health Advisor, and Mr N. Ecker-Racz, US AID Mission, Kathmandu; Dr L. Kartman, Director, Plague Labora- tory, Zoonoses Section, Fort Collins, Colorado (for- merly San Francisco, California); Mr J. Feeley, Epi- demiological Services Laboratory Section, Epidemiology Program, and Miss B. Pittman, Bacterial Chemistry Unit, Bacteriology Section, Laboratory Division, Center for Disease Control, Atlanta, Georgia, and Dr C. York, Professor of Comparative Medicine, University of Cali- fornia, San Diego, California. RtSUME OBSERVATIONS CLINIQUES ET EPIDEMIOLOGIQUES AU COURS D'UNE EPIDEMIE DE PESTE AU NEPAL Du 6 septembre au 5 novembre 1967, le village de Nawra, situe dans une region montagneuse du nord-ouest du Nepal, a connu une epidemie de peste au cours de laquelle ont e denombres 26 cas et 18 deces. Les deux premiers malades ont contracte l'infection dans un paturage, eloigne de 5 km du village, oiu l'on avait trouve des carcasses de ((rongeurs*. Des serums preleves chez trois chiens de la region contenaient des anticorps pour la fraction 1 de Pasteurella pestis. Les autres cas ont ete observes dans le village meme out la peste avait e introduite par l'un des deux premiers malades. D'apres les observations cliniques et epidemio- logiques, la transmission s'est faite d'une part par la voie aerienne, entrainant l'apparition de 6 cas d'angine pesteuse et d'un cas de peste pneumonique primaire, et, d'autre part, par l'intermediaire de puces infectees qui ont e a l'origine des 17 cas de peste bubonique. Huit des 9 malades qui ont pu etre examines etaient atteints de lesions cutanees allant de la pustule de petites dimensions a l'ulc&ration etendue. Un patient atteint de peste bubo- nique a presente des complications meningees. Six villages voisins de Nawra ont e visites au cours de I'enquete. On n'y a d&el aucun signe d'une affection ressemblant a la peste et les habitants n'avaient aucun souvenir d'une epidemie ou de cas isoles d'une maladie de ce genre. Dans le village touche, aucun indice d'une epizootie parmi les rats et autres rongeurs n'a ete decou- vert. Grace a une quarantaine stricte, a des pulverisations de DDT et a la prophylaxie par antibiotiques, on a reussi a circonscrire 1'6pid6mie. Apres analyse des donnees epidemiologiques, il est apparu que les cas de peste bubonique etaient nettement groupes. En l'absence de toute presomption d'une epizootie parmi les rongeurs, on conclut que la trans- mission de l'infection pesteuse s'est operee d'homme A homme par l'intermediaire d'ectoparasites, probablement Pulex irritans. Les deux souches de P. pestis isol&es chez des malades de Nawra acidifiaient le glycerol et reduisaient les nitrates, alors que toutes les souches de bacilles pesteux isolees en Inde sont glycerol-negatives. Selon les auteurs, le foyer nepalais n'a apparemment aucun lien avec les foyers pesteux de l'Uttar Pradesh, province indienne contigue au Nepal, et represente probablement l'extension la plus meridionale de la zone pesteuse d'Asie centrale. REFERENCES Baltazard, M. & Bahamanyar, H. (1960) Bull. Wid Hlth Org., 23, 169-215 Baltazard, M., Davis, D. H. S., Devignat, R., Girard, G., Gohar, M. A., Kartman, L., Meyer, K. F., Parker, M. T., Pollitzer, R., Prince, F. M., Quan, S. F. & Walle, P. (1956) Bull. Wid Hlth Org., 14, 457-509 Blanc, G. (1956) Rev. Hyg. Med. soc., 4, 535-562 Blanc, G. & Baltazard, M. (1945) Arch. Inst. Pasteur Maroc., 3, 173-348 Chen, T. H. & Meyer, K. F. (1966) Bull. Wld Hlth Org., 34, 911-918 Chosky, N. H. (1909) Amer. J. med. Sci., 138, 351-356 Chun, J. W. H. (1936) Clinical Features. In: Wu Lien-teh, Chun, J. W. H., Pollitzer, R. & Wu, C. Y., Plague: a manual for medical and public health workers, Shanghai, chapter 8 Francis, C. R. (1879-80) Trans. epidem. Soc. Lond., 4, 391-414 Francis, C. R. & Pearson, F. (1854) Indian Ann. med. Sci., 1, 609-645 Girard, G. (1943) Bull. Soc. Path. exot., 36, 1-43 Kartman, L., Martin, A. R., Hubbert, W. T., Collins, R. N. & Goldenberg, M. 1. (1967) Publ. Hlth Rep. (Wash.), 82, 1084-1094 2 706 F. M. LAFORCE AND OTHERS Landsborough, D. & Tunnel, N. (1947) Brit. med. J., 1,4 Machiavello, A. (1951) Plague. In: Gradwohl, R. B. H., Benitez Soto, L. & Felsenfeld, O., ed. Clinical tropical medicine, St. Louis, C. V. Mosby, p. 460 Martin, A. R., Hurtado, F. P., Plessala, R. A., Hurtado, E. G., Chapman, C. E., Callahan, E. C. & Brutsche, R. L. (1967) Pediatrics, 40, 610-616 Meyer, K. F. & Larson, A. (1960) The pathogenesis of cervical plague developing after exposure to pneumonic plague produced by intratracheal infection in primates. In: Proceedings of the Symposia held during the Diamond Jubilee of the Haffkine Institute, 1959, Bombay, pp. 1-12 Moody, M. D., Winter, C. C. (1959) J. infect. Dis., 104, 288-294 Pollitzer, R. (1954) Plague, Geneva, World Health Organization Raettig, H. (1961) Plague pandemia of the 20th century. In: Rodenwalt, E. & Jusatz, H. J., ed. World atlas ofepidemic diseases, Hamburg, Falk-Verlag Renny, C. (1851) Medical report on the mahamurree in Gurhwal in 1849-1850, Agra, cited by Francis & Pearson (1854) Rust, J. H., Cavanaugh, D. C., O'Shita, R. & Marshall, J. D. (1971) J. infect. Dis., 124, 522 Rust, J. H., Miller, B. E., Bahmanyar, M., Marshall, J. D., Purnareja, S., Cavanaugh, D. C. & U Saw Tin Hla, (1971) J. infect. Dis., 124, 527 Simpson, W. J. A treatise on plague dealing with the historical, epidemiological, clinical, therapeutic, and preventive aspects of the disease. Cambridge University Press, pp. 178-180 Swellengrebel, N. H. (1953) Docum. Med. geogr. trop. (Amst.), 5, 151-156 Wu Lien-teh (1926) A treatise on pneumonic plague, Geneva, League of Nations Publications, Health organization (CE474)