Background. Northern Thailand's biggest botulism outbreak to date occurred on 14 March 2006 and affected 209 people. Of these, 42 developed respiratory failure, and 25 of those who developed respiratory failure were referred to 9 high facility hospitals for treatment of severe respiratory failure and autonomic nervous system involvement. Among these patients, we aimed to assess the relationship between the rate of ventilator dependence and the occurrence of treatment by day 4 versus day 6 after exposure to bamboo shoots (the source of the botulism outbreak), as well as the relationship between ventilator dependence and negative inspiratory pressure.
Methods. We reviewed the circumstances and timing of symptoms following exposure. Mobile teams treated patients with botulinum antitoxin on day 4 or day 6 after exposure in Nan Hospital (Nan, Thailand). Eighteen patients (in 7 high facility hospitals) with severe respiratory failure received a low- and high-rate repetitive nerve stimulation test, and negative inspiratory pressure was measured.
Results. Within 1–65 h after exposure, 18 of the patients with severe respiratory failure had become ill. The typical clinical sequence was abdominal pain, nausea and/or vomiting, diarrhea, dysphagia and/or dysarthria, ptosis, diplopia, generalized weakness, urinary retention, and respiratory failure. Most patients exhibited fluctuating pulse and blood pressure. Repetitive nerve stimulation test showed no response in the most severe stage. In the moderately severe stage, there was a low-amplitude compound muscle action potential with a low-rate incremented/high-rate decremented response. In the early recovery phase, there was a low-amplitude compound muscle action potential with low- and high-rate incremented response. In the ventilator-weaning stage, there was a normal-amplitude compound muscle action potential. Negative inspiratory pressure variation among 14 patients undergoing weaning from mechanical ventilation was observed. Kaplan–Meier survival analysis identified a shorter period of ventilator dependency among patients receiving botulinum antitoxin on day 4 (P = .02).
Conclusions. Patients receiving botulinum antitoxin on day 4 had decreased ventilator dependency. In addition, for patients with foodborne botulism, an effective referral system and team of specialists are needed.
In northern Thailand, at least 2 outbreaks of foodborne botulism following consumption of home-canned bamboo shoots were reported. The first outbreak, affecting 4 female and 2 male patients, occurred in Mae Sot District, Tak Province, in December 1997. The second outbreak occurred in Thawangpha District, Nan Province, in April 1998. A total of 13 patients with cases of botulism were reported; of these, 9 (69.2%) were hospitalized, and 4 (30.8%) of these patients required mechanical ventilation. All patients experienced neurological symptoms, such as dysphagia, ptosis, and generalized weakness, and 4 patients had gastrointestinal symptoms, including abdominal pain, nausea, vomiting, and diarrhea [1, 2].
On 14 March 2006, the biggest botulism outbreak in northern Thailand to date affected 209 people attending a local Buddhist festival. At least 42 patients developed respiratory failure, neuromuscular failure, and autonomic nervous system failure. Among patients with mild cases, most developed only gastrointestinal symptoms or mild dysphagia, dysarthria, diarrhea, ptosis, or diplopia without neuromuscular respiratory failure. In severe cases, the main symptoms of dysphagia, dysarthria, diarrhea, ptosis, diplopia, and generalized respiratory and muscle weakness developed after ingestion of bamboo shoots preserved in a sealed bucket under anaerobic conditions; all patients consumed the shoots at the festival in Banluang District, Nan Province, Thailand [3–5].
Botulinum toxin, produced by Clostridium botulinum and taken up by peripheral cholinergic nerve terminals, inhibits acetylcholine release and subsequently causes reversible denervation of muscle fibers. After a nonsymptomatic incubation period lasting 12–36 h, nonspecific symptoms of nausea, diarrhea, ophthalmoparesis, bulbar palsy, respiratory muscle paralysis, dilated pupils, bradycardia, hypertension, and hyperhidrosis may occur. Classical electrophysiological findings in cases of botulism include low compound muscle action potential (CMAP), decremented response to repetitive nerve stimulation test (RNST) at a low stimulation frequency (3 Hz), and facilitation at high stimulation frequencies (10–50 Hz) [6].
This outbreak was a serious health threat but was promptly addressed by Thailand's Ministry of Public Health, which provided >22 million baht (∼$550,000 USD) for imported botulinum antitoxin (anti-BoTN) (∼$10,000 USD) and for referral of severely ill patients to higher-level facilities (∼$540,000 USD). Because of the restricted amount of anti-BoTN imported from the United States (50 doses from the Centers for Disease Control and Prevention; Atlanta, GA), the United Kingdom (20 doses), Canada (10 doses), and Japan (23 doses, donated by the Japanese government), only 2 lots of imported anti-BoTN were administered. Anti-BoTN was administered on either day 4 (the first lot) or day 6 (the second lot; delay of antitoxin treatment was affected by the availability of the antitoxin) after exposure to bamboo shoots on 14 March 2006, depending on the severity of the illness of patients at Nan Hospital.
Patients. We reviewed the circumstances and timing of the clinical symptoms that occurred after patients ingested pickled bamboo shoots preserved in an anaerobically sealed container. Field teams from the Ministry of Public Health treated the patients by administering anti-BoTN on day 4 or day 6, and supportive care was provided at Nan Hospital. Patients with severe respiratory failure were referred to 7 hospitals in Thailand: Rajavithi Hospital (Bangkok; 3 patients), Lumpang Hospital (Lumpang; 3 patients), Buddhachinaraj Hospital (Pitsanulok; 3 patients), Phramongkutklao Army Hospital (Bangkok; 3 patients), Maharaj Nakom Chiang Mai Hospital (Chiang Mai; 2 patients), Siriraj Hospital (Bangkok; 2 patients), and Bhumibol Hospital (Bangkok; 2 patients). All patients with severe respiratory failure were administered a low-rate (3 Hz) RNST and a high-rate (10 Hz, 20 Hz, 30 Hz, 40 Hz, and 50 Hz) RNST in the abductor digiti minimi. RNST of the musculus nasalis was not performed, because CMAP could not be elicited for most of the patients. For 14 patients, continuous measurements of negative inspiratory pressure (NIP) and maximum fractional inspiration oxygen and positive-end expiratory pressure were recorded to help the attending physician decide as expeditiously as possible whether to begin weaning from mechanical ventilation. Continuous noninvasive monitoring of pulse and blood pressure was performed to detect autonomic dysfunction. Some patients received early tracheostomy for long-term ventilator care at Nan Hospital. Examination and culture of sputum samples, Gram staining and culture of urine samples, culture of blood samples, and culture of wound samples were performed in all cases. Antibiotic treatment was dependent on the judgment of the specialty physicians at each hospital and the known susceptibility of the organism to a given antibiotic.
Statistical analysis. For continuous data, mean values ± SDs and 95% CIs were used as descriptive statistics. We calculated the mean h of onset and 95% CI following ingestion of bamboo shoots. Our estimate was that all patients consumed the C. botulinum spore and/or botulinum toxin at noon on 14 March 2006. The Kaplan-Meier survival model was used to calculate the cumulative rate of ventilation dependence between the 2 groups of patients treated with anti- BoTN on day 4 or day 6. Pearson's correlations for worst NIP, the effect of NIP on the success of weaning from mechanical ventilation, and duration of dependence on mechanical ventilation were calculated.
All 18 patients with severe neuromuscular respiratory failure (NIP, <15 mm Hg) became ill 1–65 h after ingestion of pickled bamboo shoots, with a mean time to onset (±SD) of 21.3 ± 16.5 h. Three patients (16.6%) had symptoms indicative of food poisoning (abdominal pain, nausea, and vomiting without diarrhea) within the first day, and 14 patients (77.7%) exhibited a delayed onset of initial symptoms (>24 h); 1 patient did not experience symptoms of food poisoning, but dizziness occurred within 2 h after ingestion. Abdominal pain occurred in 12 patients (66.6%), with onset occurring 1–50 h after ingestion of bamboo shoots, with a mean time to onset (±SD) of 19.0 ± 4.2 h. Most patients (83.3%) experienced nausea and vomiting, with a time to onset of symptoms of 1–65 h and a mean time to onset (±SD) of 22.8 ± 4.2 h. Onset of diarrhea occurred 10–50 h after ingestion of bamboo shoots, with a mean time to onset (±SD) of 26.1 ± 5.1 h.
Neuromuscular symptoms manifested as dysphagia and/or dysarthria (mean time to onset [±SD], 37.0 ± 3.9 h; range, 21–76 h), ptosis (mean time to onset [±SD], 41.6 ± 5.1 h; range, 12–90 h), generalized weakness (mean time to onset [±SD], 45.5 ± 4.9 h; range, 19–84 h), and respiratory failure (mean time to onset [±SD], 72.4 ± 4.1 h; range, 46–97 h). Neurological physical examination identified ptosis and/or ophthalmoparesis in 16 (88.8%) of the patients, proximal muscle weakness in 13 (72.2%), paradoxical abdominal respiration in 18 (100%), hyporeflexia in 10 (55.5%), and enlarged pupils nonreactive to light in 3 (16.6%).
Interestingly, autonomic involvement, including fluctuation of pulse and blood pressure (100%), and urinary retention (72.2%) were present in most patients (table 1). Cardiovascular complications consisted of transient sinus bradycardia in 6 patients (33.3%) and atrial fibrillation in 1 patient (5.5%). The progression of symptom onset seems to be a typical sequence (figure 1).
Onset of symptoms of botulism in patients with severe neuromuscular respiratory failure after exposure to contaminated bamboo shoots. Time of exposure was defined as 12 P.M. on 14 March 2006. Severe neuromuscular respiratory failure was defined as occuring in patients requiring mechanical ventilation with unstable blood pressure and/or pulse. Gray bars indicate range of values, black bars indicate median values, and brackets indicate 95% CIs. Outlier values are indicated by stars or circles.
The pattern of RNST consisted of no response elicited from patients with the most severe stage of botulism (defined as complete respiratory paralysis from severe respiratory muscle weakness and NIP <15 mm Hg with mechanical ventilation), especially in the musculus nasalis, which was more refractory than the abductor digiti minimi; most patients at this stage had to be stimulated at a high voltage (>160 mV) to elicit a response. In patients with a moderately severe stage of botulism (defined as incomplete respiratory paralysis and NIP >15 mm Hg with mechanical ventilation), RNST-elicited CMAP showed initially very low amplitudes; low-rate RNST (3 Hz) elicited an incremented response (figure 2A), whereas high-rate RNST (10 Hz) elicited a decremented response (figure 2B). For patients in the early recovery stage (defined as improved respiratory muscle weakness, improved sequential NIP, and the ability to move all extremities), RNST-elicited CMAP showed initially low amplitudes, with incremented response with both low-rate (3 Hz) and high-rate (10 Hz, 20 Hz, 30 Hz, 40 Hz, and 50 Hz) stimulation (figure 2C). In the ventilator-weaning stage, RNST produced nearly normal or normal CMAP amplitude with normal or mild incremented response with both low-rate (3 Hz) and high-rate (50 Hz) stimulation (figure 2D). The recovery pattern was correlated with overall motor power and diaphragm functions. NIP variation among patients on the day on which mechanical ventilation was removed (i.e., the extubation day) was observed. There was a strong, statistically significant positive correlation between worst NIP and the duration of mechanical ventilation (Pearson's correlation, 0.772; P = .042). However, there was no significant correlation between NIP on the extubation day and the duration of mechanical ventilation (Pearson's correlation, -0.333; P = .347). The variation in NIP on the extubation day ranged from -12 to -51 mm Hg, with a mean NIP (±SD) of -26.5 ± 11.7 mm Hg.
Results of repetitive nerve stimulation tests of the abductor digiti minimi. Stimulation at 3 Hz (A) elicited an incremented response (white arrow) in the moderately severe stage of botulism (defined as occurring in patients with incomplete paralysis of respiratory muscles). Stimulation at 10 Hz (B) elicited a decremented response (white arrow) in the moderately severe stage of botulism. Stimulation at 50 Hz (C) elicited an incremented response (white arrow) in both the moderately severe stage and the early recovery stage of botulism. Stimulation at 3 Hz after complete recovery from botulism (D) elicited a normal response (white arrow).
Complications in the intensive care unit included aspiration pneumonia (in 83.3% of patients), acute pulmonary edema (11.1%), acute respiratory distress syndrome (5.5%), urinary tract infection (27.7%), and tracheostomy tract infection (5.5%). The identified organisms from sputum samples obtained from patients with aspiration pneumonia were Enterococci species (1 isolate), Escherichia coli (1 isolate), and Klebsiella pneumoniae (3 isolates). The organisms obtained from patients with nosocomial pneumonia were Acinetobacter baumannii (2 isolates), Pseudomonas aeruginosa (7 isolates), and Staphylococcus aureus (2 isolates). From patients with nosocomial urinary tract infection, isolated strains were Pseudomonas aeruginosa (2 isolates), Enterococci species (1 isolate), and Candida albicans (1 isolate). From the patient with nosocomial septicemia, the isolated strain was Burkholderia ceptacia (1 isolate), and from patients with wound infection, the isolated strains were Enterobacter speices (1 isolate) and Klebsiella species (1 isolate).
Kaplan-Meier survival analysis showed no statistically significant difference in cumulative rate of respiratory failure (time to intubation after exposure to bamboo shoots on 14 March 2006) between patients with anti-BoTN administration on day 4 and those with anti-BoTN administration on day 6 (P = .27) (figure 3). Kaplan-Meier survival analysis showed a statistically significant difference with respect to duration of the period from exposure to bamboo shoots to extubation day between patients who received treatment with anti-BoTN on day 4 and those who received treatment on day 6 (P = .022) (figure 4). The patients who were treated on day 4 had a reduced duration of mechanical ventilation (P = .028) (figure 5). Finally, all patients survived after receiving treatment with anti-BoTN.
Cumulative rate of respiratory failure for 18 patients treated with botulinum antitoxin (anti-BoTN) on day 4 or day 6 after ingestion of contaminated bamboo shoots. Time to onset of respiratory failure is defined as the time period from exposure to contaminated bamboo shoots (defined as 12 P.M. on 14 March 2006) to intubation for mechanical ventilation. Kaplan-Meier survival analysis showed no statistically significant difference with respect to time to onset of respiratory failure between patients who received anti-BoTN on day 4 after exposure and patients who received anti-BoTN on day 6 (P = .27).
Duration of the period between exposure to contaminated bamboo shoots and extubation after mechanical ventilation for 18 patients treated with botulinum antitoxin on day 4 (day 4 group) or day 6 (day 6 group) after exposure. The date of exposure to contaminated bamboo shoots was defined as 14 March 2006. Kaplan-Meier survival analysis showed a statistically significant difference between the day 4 group and the day 6 group with respect to the duration of this period (P = .02).
Duration of mechanical ventilation for 18 patients treated with botulinum antitoxin on day 4 (day 4 group) or day 6 (day 6 group) after exposure to contaminated bamboo shoots on 14 March 2006. Kaplan-Meier survival analysis showed a statistically significant difference between the day 4 group and the day 6 group with respect to the duration of mechanical ventilation (P = .02).
Foodborne botulism in northern Thailand is rarely reported and usually occurs in small numbers of people after ingestion of home-canned bamboo shoots [1, 2]. In a previous outbreak, 4 (30.7%) of 13 infected people required mechanical ventilation. None of the patients associated with that outbreak received treatment with anti-BoTN, and 2 (15.4%) of 13 patients died. In the outbreak analyzed here, 42 (20.1%) of 209 patients required mechanical ventilation, and all of these patients received an infusion of anti-BoTN on either day 4 or day 6 after ingesting contaminated bamboo shoots. There were no reports of death associated with this outbreak. A total of 133 (63.6%) of 209 patients were female; the higher percentage of female patients is probably attributable to the fact that the contaminated bamboo shoots were a local product that was made by female housemate groups in the village. Some patients reported that the taste of the contaminated bamboo shoots had been a slightly sweeter than usual.
A total of 17 patients who were admitted to Nan Hospital with botulism during this outbreak were not included in the current analysis, because they did not complete the RNST protocol before weaning from mechanical ventilation. A total of 18 (72.0%) of the 25 patients who had a more severe clinical status and were referred to 5 hospitals completed the RNST protocol before weaning from mechanical ventilation. However, 2 patients who were referred to Siriraj Hospital and 2 patients who were referred to Bhumibol Hospital did not complete the RNST protocol before weaning from mechanical ventilation. Fourteen patients (77.7%) were weaned from mechanical ventilation as quickly as possible after RNST results indicated regression of abnormalities, NIP improved, and overall clinical motor power was recovered.
One (5.5%) of the 18 patients who completed the RNST protocol before weaning from mechanical ventilation had to undergo reintubation; this occurred because the patient had inspiratory stridor resulting from vocal cord edema without signs of paradoxical abdominal respiration or muscle weakness and without laboratory indications of hypokalemia. In the patients, the clinical manifestations of botulism developed in the typical sequence, as shown in figure 1. The 3 patients (16.6%) who exhibited symptoms characteristic of food poisoning and the 15 patients (83.3%) who did not had the same time to onset of respiratory failure (table 1). This finding suggests that all of these patients were exposed to spores of C. botulinum that required the same incubation period before inducing severe respiratory failure and that the food-poisoning symptoms may have been related only to high amounts of the toxin itself [7–9]. Moreover, this hypothesis regarding C. botulinum spore incubation was confirmed by culture of C. botulinum from samples of the bamboo shoots remaining in the preservative container; botulinum toxin type A and type non-A non-B were identified.
Demographic, clinical, and laboratory characteristics of 18 patients involved in an outbreak of botulism in Thailand.
The motor weakness of patients with severe respiratory failure seemed to be more predominant in the diaphragm, intercostal muscle, and facial muscles. The effects of this motor weakness caused the major complications in these patients, such as aspiration pneumonia, nosocomial pneumonia, and lung atelectasis. However, idiopathic pulmonary edema and acute respiratory distress syndrome occurred more frequently than other complications, especially among patients with volume overload and rapid response to treatment with furosemide injection and volume restriction. When treating patients with furosemide injection and volume restriction, it is important to exercise caution during the hypotensive period and to use an infusion of fluid, such as normal saline.
Cardiovascular involvement, predominantly bradycardia, may be the result of the toxin itself, and in our patients, recovery occurred 1–2 weeks after treatment with anti-BoTN. Both cardiovascular recovery and recovery from generalized muscle weakness seemed to occur more rapidly than recovery of the facial and diaphragm. Almost 16 patients had nearly full motor power in their extremities before weaning from mechanical ventilation, and diaphragm and ocular muscle recovery occurred at approximately the same time.
RNST of the abductor digiti minimi muscle could be performed more easily than RNST of other muscles, because the recovery of the nasalis and orbicularis oculi muscles was very slow and occurred only after successful weaning from mechanical ventilation. The parameters that best predicted successful weaning from mechanical ventilation were recovery of CMAP to normal amplitudes and normal RNST results (i.e., RNST results with no incremented or decremented responses) with stimulation at a low frequency (3 Hz), because NIP varied among the patients. Some pulmonologists and intensivists noted variation in NIP in individual patients within the same day. Worst NIP was a predictor of duration of dependence on mechanical ventilation, but best NIP did not predict successful extubation, and some patients underwent successful extubation with low NIPs (e.g., -12 to -15 mm Hg). Therefore, we concluded that RNST results are a useful parameter for guiding weaning from mechanical ventilation among botulism patients.
This report, which is, to our knowledge, the clearest quantified documentation to date of the effectiveness of equine-source botulinum antitoxin, confirms the idea that the infusion of anti-BoTN as quickly as possible enhances favorable outcomes however, this is, to our knowledge, the first measurable demonstration that the cumulative rate of dependence on mechanical ventilation can be significantly reduced when treatment occurs by day 4 after exposure to botulinum toxin (which, in this outbreak, had an incubation period of <40 h) (figure 1) [1, 2, 8–13].
In conclusion, this accidental botulism outbreak also demonstrated that, with this infection, many patients require more than specific anti-BoTN treatment within 4 days of exposure to experience a favorable outcome. A large botulism outbreak can severely tax many health care facilities, and it is clear that a readily available supply and rapid administration of anti-BoTN would prevent substantial morbidity in the event of future outbreaks in Thailand. Moreover, successful treatment of severely affected patients required the expertise of a neurologist, a pulmonologist, an intensivist, a cardiologist, and infectious diseases specialists, as well as good rehabilitation and referral services.
Wilai Puavilai, Piraj Kateruttanakul, Subsai Kongsaengdao, Poj Intalapaporn, Petcharat Dusitanond, Piyathida Harnsomburana, Iyavut Thaipisuttikul, Worapojn Laowittawas, and Chumpita Sutthapas (Rajavithi Hospital); Kanoksri Samintarapanya, Uthai Phengtham, Kanlaya Panjpornpon, and Aran Rattanaphon (Lumpang Hospital); Siwarit Rusmeechan, Jirayut Janma, Jakapong Lorsuwansiri, and Naritchaya Phattana (Buddhachinaraj Hospital); Adisorn Wongsa, Kunchit Piyavechviratana, Pasiri Sithinamsuwan, Jithanorm Suwantamee, Wanna Wongmek, and Parnsiri Chairangsaris (Phramongkutklao Army Hospital); Chaicharn Pothirat, Chalerm Liwsrisakun, Chaiwat Bumroongkit, Athavudh Deesomchok, and Theerakorn Theerakittikul (Maharaj Nakorn Chiang Mai Hospital); Chairat Permpikul, Surat Tongyoo, Warakarn Vilaichone, Ranistha Ratanarat, Akekarinth Poompichate, and Kanokwan Boonyapisit (Siriraj Hospital); Sunsanee Pongpakdee, Saengduan Mayotarn, and Benjamas Piya-Isragul (Bhumibol Hospital); Hathai Panyadilok (Banluang Hospital); and Niwatchai Sutcharitchan, Apinya Chuesuwan, and Petchdee Oranrigsupau (Nan Hospital).
We thank the Public Health Minister and Deputy Public Health Minister of Thailand; Dr. Prat Boonyavongvirot, Permanent Secretary of the Public Health Ministry of Thailand; Dr. Chatri Banchuin, Director-General of the Department of Medical Services of Thailand; the house staff and director of Nan Hospital; the mobile teams from the Ministry of Public Health, Thailand; Dr. Jedsada Chokdamrongsuk, Director of Rajavithi Hospital; the medical volunteers from Ramathibodi Hospital and Chulalonkorn Hospital; Dr. Samart Nidhinandana; Dr. Yotin Chinvarun, from Phramonkutklao Army Hospital; Dr. Kongkiat Kulkantrakorn, from Thammasart Hospital; and Assistant Professor Dusit Sujirarat, from Mahidol University, for statistical guidance.
Financial support. Rajavithi Hospital, Department of Medical Service, Public Helath Ministry of Thailand (RVH_CER_001). Botulinum antitoxin was provided by the Ministry of Public Health, Thailand.
Potential conflicts of interest. All authors: no conflicts.
IDSA Members: For your free access to this journal, log in via the IDSA members area.
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy