Abstract
Abstract
Background
Febrile illnesses of bacterial origin, including enteric fever, leptospirosis, murine typhus, and scrub typhus are potentially fatal and prevalent in Asia. The standard diagnosis test including microscopic agglutination test (MAT) for leptospirosis and immunofluorescent assay (IFA) for rickettsiosis are not available in Myanmar. As an results, these diseases are under diagnosis and misdiagnosis with other febrile illness. The rapid alternative diagnosis test like polymerase chain reaction (PCR) would be invaluable to improve the diagnosis and recognition of these infections in Myanmar. In addition, the use of antimicrobials before collection of samples for diagnostic microbiology investigations can result in a substantially lower diagnostic sensitivity through bacterial killing. Moreover, increased prevalence of antimicrobial resistance among common pathogens is major concern and pre-hospitals antimicrobials use may have impact on diagnostic tests sensitivity. To provide baseline information of antimicrobial use and resistance at Yangon General Hospital (YGH), we planned to do a Global point prevalence surveys (Global-PPS) at that facility to fill gaps in information about antimicrobial consumption and resistance at YGH, and provide the evidence base for developing locally relevant, updated antimicrobial guidelines. The typhoid conjugate vaccine (TCV) have been introduce in neghibouring countries and improved data on typhoid fever epidemiology are needed to inform country-level decisions about typhoid conjugate vaccine introduction and strategies in Myanmar.
Objectives
This PhD project was designed to evaluate the urine rtPCR for diagnosis of leptospirosis and rickettsiosis, to detect the pre-hospital antimicrobial use and its impact on diagnostic tests sensitivity, to understanding antimicrobial consumption and resistance at Yangon General Hospital (YGH), and to estimate the incidence of enteric fever, leptospirosis, rickettsiosis in Yangon, Myanmar.
Methods
Three projects were included in this study. Firstly, we did two systemic reviews, one to overview the urine PCR for systemic infectious diseases and then did another systemic review to choose the apporiate targeted for urine rtPCR to detect the leptospirosis and rickettsiosis in our community onset febrile study.
We nested our study within a prospective cohort study of community onset febrile illness at YGH. As described in detail previously, patients ≥12 years old with fever ≥38˚C who attended YGH from 5 October 2015 through 4 October 2016 were recruited. We performed blood culture, detection of Leptospira by microscopic agglutination test (MAT), detection of murine typhus, scrub typhus, spotted fever rickettsiosis by immunofluorescence antibody assay (IFA), and rtPCR in haemoculture fluid (HCF) for Leptospira and Rickettsia. Isolation, identification, and antimicrobial susceptibility testing of organisms were done for all positive blood culture bottles. To confirm Salmonella serovars, whole genome sequencing was performed. The MAT panels including suspensions of 22 living cultures representing 22 serovars and 20 serogroups of Leptospira were used to detect antibodies of Leptospira in patients’ sera. IFA slides were used for detection of antibodies to Orientia tsutsugamushi pooled Karp, Kato, Gilliam antigens and Rickettsia typhi Wilmington strain in participants sera.
The participants who provide urine were enrolled in our study, we detected urine antimicrobial activity by modified Kirby-Bauer method using three reference organisms; Bacillus subtilis American Type Culture Collection (ATCC) 6633, Escherichia coli (ATCC 25922), and Streptococcus pyogenes (ATCC 19615). We performed rtPCR in urine to detection Leptospira species in urine by rrs16S rRNA rtPCR and lipL32 rtPCR, Rickettsia species by genus specific 17kDa antigen rtPCR and species-specific assay ompB R. typhi rtPCR and O. tsutsugamushi by 47kDa (htrA) rtPCR.
Secondly, we did healthcare utilization survey (HCUS) at Yangon Region based on two stage cluster survey methodology. In first stage, 48 wards were randomly selected from 689 wards and 336 households were selected in second stage. The participants were asked and answered separately about usual healthcare seeking behaviour in the course of fever ≥3 days by different age group, together with actual healthcare seeking behaviour of every household members who suffered from fever in the past 3 months. We estimated incidence with the use of multipliers derived from the healthcare utilization survey and the community onset febrile illness study. Multipliers accounted for persons with enteric fever, leptospirosis, murine typhus, scrub typhus, and spotted fever ricketttsiosis who would potentially be missed through the stages of reporting including healthcare facility choice, referral from another inpatient’s facility, and diagnostic test sensitivity. The multiplicative inverse of the relevant proportions are multipliers.
In third project, we used the standardised Global Point Prevalence Survey (Global PPS) methodology to conduct survey at YGH. We conducted the survey to all inpatients on any ward of YGH at 8:00 am on the day of data collection were included in survey. All inpatients were counted and the antimicrobials excluding topical antibiotics, prescribed to any patients were recorded in detailed. All inpatients counted towards the patient denominator. Antimicrobial use, excluding topical antimicrobials, prescribed to any patient contributed to the numerator Antimicrobials were recorded by generic name. We also recorded data on patient characteristics, biomarker information, and microbiology results from patient medical records.
Results
According to our two systemic review, we found that limited role in diagnosis of diseases like legionellosis, tuberculosis, leptospirosis, but has a poor diagnostic test for other infection. Based on our reviewed finding we chose rrs 16S rRNA and lipl32 rtPCR assay combination for diagnosis of leptospirosis, the 47kDa gene rtPCR was for diagnosis of scrub typhus and the 17 kDa genus specific gene rtPCR together with Rickettsia typhi specific ompB rtPCR assay combination for diagnosis of rickettsia infection in our urine specimens from community onset febrile illness study.
Leptospira species was detected in 6.3% of participants by urine rrs 16S rRNA and 0.2% by urine lipL32 rtPCR. The sensitivity and specificity of urine lipL32 rtPCR against serologically confirmed and probable leptospirosis by MAT was 7.7% and 100% whereas sensitivity and specificity of rrs 16S rRNA rtPCR was 11.5% and 93.9%, respectively. Only one participants (0.1%) was positive for 47-kDa rtPCR in urine and none for rickettsia infection by urine rtPCR.
The urine antimicrobials activity was detected in 80.8% of febrile participants and the sensitivity and specificity of reported antibiotic use compared to urine antimicrobial activity detection were 37.8% and 75.2%. Urine antimicrobial activity was not associated with the odds of a pathogen being isolated from blood culture (odds ratio 0.97, 95% CI 0.5% - 1.9%, P = 0.941) but it had impact on detection of Rickettsia species by 17kDa urine rtPCR (odds ratio 0.26, 95% CI 0.10% - 0.71%, P = 0.001).
We found that at least one antimicrobial was prescribed to 64.4 % of admitted patients in YGH with third generation cephalosporin accounted 33.3% of all antibacterial and most frequent diagnosis for antimicrobial prescriptions was pneumonia (16.0%). Only 15% an antimicrobial prescription in YGH were informed by a culture susceptibility result. The surgical prophylaxis accounts for 75.1% for prophylaxis indication and 77.6% of which were given more than one day. Guidelines were not available for 59.8% of indications.
Of 671 febrile participants from Yangon Region from community onset febrile illness study at YGH, 17 had confirmed leptospirosis, and seven had probable leptospirosis. Scrub typhus was positive by IFA serology in eight cases, murine typhus in three cases, and there were no cases of spotted fever rickettsiosis. Of blood culture positive cases, 42 were confirmed to have enteric fever with 33 cases of typhoid fever and nine cases of paratyphoid fever.
We estimated the incidence of diseases by using mutlipliers derived from HCUS and data from community onset febrile illness study. We estimated that the annual incidence of leptospirosis among adolescents and adults as 158 – 239 per 100,000 population, whereas incidence of murine typhus among adolescents and adults as 44 per 100,000 population, and overall annual scrub typhus incidence among adolescents and adults as 219 cases per 100,000 persons year for 2015–16 in the Yangon Region. We estimated that annual incidence of enteric fever among adolescents and adults was 498 per 100,000 population, with typhoid incidence in this age group 391 per 100,000 persons and paratyphoid incidence 107 per 100,000 persons, in 2015-2016 in Yangon Region.
Conclusion
We found that urine PCR lacks sensitivity for the diagnosis of leptospirosis, rickettsiosis, and scrub typhus. Although the diagnosis of leptospirosis by urine PCR is possible, but PCR is not sufficiently sensitive to contribute to the diagnosis of rickettsiosis and scrub typhus. Moreover, urine antimicrobial activity was detected in signifant proportion of febrile participants and participants medicine use history were unreliable. We found that prior antimicrobial use did not have significant effect on blood culture sensitivity but it has effect on detection of Rickettsia species by rtPCR. We found that high prevalence of third generation cephalosporin and fluoroquinolones prescription in YGH and metronidazole was mostly used for prophylaxis. Our findings point that a need for community antimicrobial stewardship (AMS) efforts as well as hospital antimicrobials stewardship programs (ASPs) while maintaining access to critical medicines through expansion of affordable health services. The incidence estimates of enteric fever, leptospirosis, murine typhus and scrub typhus are high. More research is needed to describe the burden of these diseases in terms of not only incidence but also disability and mortality as well as sources, and risk factors for human infections. Typhoid conjugate vaccines offer long-term protection in infants and children and thus should be considered in Myanmar.