Risk of Tuberculosis Infection among Healthcare Workers in a Tertiary Care Hospital in Bengaluru City.

INTRODUCTION: Tuberculosis (TB) is an airborne communicable disease caused by Mycobacterium tuberculosis. Infection is most likely to occur when an individual is exposed to someone with pulmonary TB on a day-to-day basis, such as by living or working in close quarters with someone who has an active disease. This study was undertaken using reliable screening and investigation tools for assessing the incident latent tuberculosis infection (LTBI) among healthcare workers (HCWs).
METHODOLOGY: This descriptive longitudinal study was conducted among 600 HCWs at a tertiary care hospital in Bengaluru city for a period of 1 year 4 months (March 2016-June 2017). All HCWs were stratified according to their work profile as doctors, nurses, laboratory technicians, nursing aides, pharmacists, laundry workers (only those who come in contact with patients), dietary workers (only those who come in contact with patients), and workers from housekeeping department. Study subjects were enrolled in proportion to their population size. The data were collected by administering TB screening questionnaire and performing tuberculin skin test (TST). All the subjects who had negative TST reaction were followed up after 1 year from March 2017 to June 2017.
RESULTS: The study subjects consisted of 600 HCWs with a mean age of 27.8 ± 7.1 years. Initially, 600 study subjects underwent TST, 2 of whom were not contactable between 48 and 72 h of test administration. Of the remaining 598 subjects, 120 (20.1%) had a positive TST reaction. After 1 year, of the 478 subjects who tested negative, 345 study subjects underwent the repeat TST and 67 (19.4%) of the study subjects had a positive TST reaction.
CONCLUSION: The prevalence of LTBI was 20.1% and it was significantly associated with age, place of residence, education, work commute time, and mode of transport. The incidence of LTBI was 19.4% and it was significantly associated with gender, place of residence, education, work commute time, and marital status. Copyright:

INTRODUCTION

Pulmonary tuberculosis (PTB) is an airborne communicable disease caused by Mycobacterium tuberculosis (Mtb). The TB bacteria are expelled in the air when a person with TB disease coughs, sneezes, speaks, or sings. The people nearby may breathe in the droplets containing Mtb and become infected. Two billion people, equal to one-third of the world's total population, are infected with TB bacilli (Mtb). One in 10 people infected with TB bacilli will become sick with active TB. Infection is most likely to occur when an individual is exposed to someone with TB disease on a day-to-day basis, such as by living or working in close quarters with someone who has the active disease. Even then, because the bacteria generally stay latent (inactive) after they invade the body, only a small number of people infected with TB will ever have the active disease. It is suggested that “because these latent infections can eventually become active, even people without symptoms should receive medical treatment. Medication helps to get rid of the inactive bacteria before they become active.”[1] HCWs are at an increased risk of infection with Mtb because they are exposed to patients with active disease in settings that favor transmission, especially when working in ill-ventilated spaces and when performing procedures involving contaminated aerosols. The risk of transmission of Mtb from individuals with TB to other healthcare workers (HCWs) has been recognized for many years. This risk is greater when large numbers of infectious (smear-positive) patients with TB are managed at a healthcare facility, and the risk of transmission can be reduced with effective infection-control measures. Screening for TB infection is recommended among HCWs who work in proximity with patients with TB.[2] In Low-Middle Income Countries which are resource-poor, implementing even low-cost strategies to reduce TB transmission among HCWs is a challenge. There are very few studies done on latent tuberculosis infection (LTBI) among HCWs in this country. India being one of the middle-income countries with a high burden of TB and with HCWs in the country being at an increased risk of getting infected with TB, there is a need to assess risk of TB infection among them. This study was undertaken using reliable screening and investigation tools for assessing the incidence of LTBI among HCWs. The results of this study will help implement preventive measures among HCWs.

METHODOLOGY

This descriptive longitudinal study was conducted among 600 HCWs at tertiary care hospital in Bengaluru city for a period of 1 year 4 months (March 2016–June 2017). After obtaining Institutional Ethics Committee clearance and permission from the Associate Director of the hospital and the Heads of the respective Departments, HCWs were recruited for the study based on the eligibility criteria. The HCWs were stratified according to their work profile as doctors, nurses, laboratory technicians, nursing aides, pharmacists, laundry workers (only those who come in contact with patients), dietary workers (only those who come in contact with patients), and workers from housekeeping department. The required sample was randomly selected from each group except doctors. The study subjects were enrolled in proportion to their population size. Due to a high nonresponse rate among the doctors, all doctors who agreed for the tuberculin skin test (TST) were interviewed and subsequently tested either once or twice depending on their reactivity. This sample was dependent on the number of uninfected persons available; the sampling was stopped on reaching the required number of uninfected persons (from each category) at initial enrolment. HCWs who gave positive TST, had TB in past, HCWs who will not be working for more than 1 year in the hospital from their date of enrolment into this study, known reported cases of HIV, individuals on long-term steroid treatment, history of anaphylactic reaction to tuberculin, HCWs who had positive TST in the past, and pregnant women were excluded from the study. Written informed consent was obtained from the HCWs before the interview. The data were collected by administering TB screening questionnaire. The LTBI screening questionnaire is one of the most widely used and validated questionnaires to screen for LTBI among HCWs in the United States. We used this questionnaire for screening HCWs as there were no standard questionnaires available in India. LTBI screening questionnaire had 14 items and an answer for each question had “yes,” “no,” or “do not know” options. Separate written informed consent was obtained for performing the TST. Following this, each study participant was administered the TST. The TST is performed by injecting 0.1 mL of tuberculin purified protein derivative (PPD) into the inner surface of the forearm. The TST is an intradermal injection. When the tuberculin is delivered correctly, the injection should produce a pale elevation of the skin (a wheal) 6–10 mm. Each subject was reviewed by the investigator between 48 and 72 h of the TST to read the reaction. All the subjects who gave negative TST reaction were followed up after 1 year. At this time, the history of exposure to patients with TB in the past 1 year was recorded and a follow-up TST was done on all consenting subjects. Again, the TST reaction was read between 48 and 72 h.

RESULT

The study subjects consisted of 600 HCWs with a mean age of 27.8 ± 7.1 years. The majority of the study subjects (538, 89.7%) were in the age group of 21–40 years and most of them were females (457, 76.2%). Most of the study subjects (412, 68.7%) were unmarried. The majority of the study subjects were Christian (387, 64.5%). More than half of the study subjects (353, 58.8%) were staying in hostels (in institutional campus). Exactly half (300, 50%) of the study subjects had completed undergraduation followed by 124 (20.7%) who had completed their postgraduation. The majority of the study subjects belonged to the upper class (392, 65.3%), followed by upper middle class (104, 17.3%) according to modified BG Prasad socioeconomic status classification. The study subjects consisted of doctors (99), nurses (277), laboratory technicians (41), nursing aides (86), pharmacy workers (24), laundry workers who come in direct contact with the patients (18), dietary workers who come in direct contact with the patients (12), and housekeeping workers (43) [Table 1]. The majority of the study subjects were nurses (277, 46.2%). More than half of the subjects (320, 53.3%) were able to reach the hospital within 10 min from their residence and 158 (26.3%) needed more than 30 min to reach the hospital from their residence. The majority of the study subjects (373, 62.2%) reached hospital by walking and 148 (24.7%) used public transport. The majority of the study subjects (535, 90.4%) had worked less than 10 years in this hospital and 45 (7.6%) of the study subjects worked for 11–20 years. The majority of the study subjects (517, 86.3%) worked less than 8 h per day. Seventy (11.7%) study subjects worked for 9–16 h and 12 (2%) worked more than 16 h. The majority of the study subjects (495, 82.5%) worked 6 days per week. Eighty-eight (14.7%) study subjects worked for all 7 days. One hundred (16.7%) of the study subjects had received some training on occupational safety. Initially, 600 study subjects underwent TST, and 2 of them were not contactable between 48 and 72 h of test administration. Of the 598 remaining, 120 (20.1%) had a positive TST reaction and 478 (79.6%) had a negative TST reaction [Figure 1]. In two subjects, TST was not read as they were not available after 48 h to read the reaction. There was a significant association between age, place of residence, education, time taken to reach hospital, and mode of transport with TST reaction [Table 2a]. Laundry workers, housekeeping staff, and pharmacy workers had a greater risk of developing TB infection [Table 2b]. People with no BCG scar had more positive TST reaction [Table 2b]. Multivariate linear regression was performed taking prevalence of TB infection as the dependent variable and all other factors were found to be statistically significant in univariate analysis as the independent variable. No single variable was found to be significant by this procedure [Table 3]. After 1 year, of the 478 subjects who tested negative, 345 study subjects underwent the repeat TST. Of these, 67 (19.4%) of the study subjects had a positive TST reaction and 278 (80.6%) of the study subjects had a negative TST reaction [Figure 1]. Two study subjects who were TST-negative at baseline had developed active TB disease in the intervening period. There is a significant association between gender, place of residence, education, time taken to reach hospital, and marital status with TST reaction. Male gender, those who have reported to be residing at home, lower educational status, longer work commute time, and married persons had a higher risk of developing TST positive reaction. The greater the duration of employment (work) at the hospital, the more the chance of acquiring TB infection. Subjects working in laundry (those who come in contact with patients), housekeeping, and pharmacy departments had a greater chance of infection [Table 4]. Multivariate linear regression was performed taking incidence of TB infection as the dependent variable and all other factors were found to be statistically significant in univariate analysis as independent variable. No single variable was found to be significant by this procedure [Table 5].

Table 1

Recruitment of study subjects based on proportion to population size

Work profile	Total employees	Required TST- negative subjects (proportion to population size)	Underwent initial TST	Underwent follow-up TST
Doctors	408	66	99	69
Nurses	857	139	277	154
Laboratory technicians	89	14	41	27
Nursing aides	303	49	86	54
Pharmacy workers	70	11	24	9
Laundry workers	41	7	18	8
Dietary workers	25	4	12	6
House keeping	89	14	43	18
Total	1882	304	600	345

TST: tuberculin skin test

Figure 1

Flow chart of the study

Table 2a

Association between sociodemographic characteristics and TB infection at baseline among the study subjects (n=598)

Variables (%)		TST-positive	TST-negative	P
Age (years)	<20	3 (12.5)	21 (87.5)	0.001*
	21-40	101 (18.8)	435 (81.2)	(Chi-square value=10.48)
	41-60	16 (42.1)	22 (57.9)
Present place of residence	Home resident	71 (28.9)	175 (71.1)	0.001*
	Hostel resident	49 (13.9)	303 (86.1)	(Chi-square value=20.1)
Education	Illiterate	7 (43.8)	9 (56.2)	0.001*
	Primary	16 (47.1)	18 (52.9)	(Fisher’s exact
	Secondary	31 (24.8)	94 (75.2)	Test value=25.1)
	Graduate	66 (15.6)	357 (84.4)

TST: tuberculin skin test; TB: tuberculosis

Table 2b

Association between occupation characteristics/BCG vaccine characteristics and TB infection at baseline among the study subjects (n=598)

Variables		TST-positive	TST-negative	P
Work commute time	<10 min	43 (13.5)	276 (85.5)	0.001*
	11-30 min	36 (29.5)	86 (70.5)	(Chi-square value=18.9)
	More than 30 min	41 (26.1)	116 (73.9)
Mode of transport	Other	66 (17.8)	305 (82.2)	0.032*
	Private	14 (16.2)	67 (83.8)	(Pearson’s Chi square=6.845)
	Public	40 (27.4)	106 (72.6)
Taken BCG vaccine	Yes	108 (21.6)	391 (78.4)	0.031*
	No	12 (12.1%)	87 (87.9)	(Chi square value=4.6)
Presence of BCG scar	Yes	97 (22.7)	330 (77.3)	0.013*
	No	23 (13.5)	148 (86.5)	(Chi square value=6.5)
	Doctors	18 (18.2)	81 (81.8)	0.001*, (Chi-square test value=27)
	Nurses	38 (13.8)	238 (86.2)
Work profile	Laboratory technicians	9 (22)	32 (78)
	Nursing aides	16 (18.6)	69 (81.2)
	Pharmacy	7 (29.2)	17 (70.8)
	Laundry	9 (50)	9 (50)
	Dietary	3 (25)	9 (75)
	Housekeeping	20 (46.5)	23 (53.3)

TB: Tuberculosis

Table 3

Multivariate analysis for prevalence of TB infection among the study subjects (n=600)

Variables	Multivariate analysis (P)	OR (odds ratio)	95% CI for OR
			Lower	Upper
Age	0.149	1.570	0.851	2.897
Place of residence	0.055	0.524	0.271	1.013
Education	0.546	0.853	0.508	1.431
Work commute time	0.577	1.114	0.761	1.632
Designation	0.475	1.082	0.871	1.345
Mode of transport	0.256	1.233	0.859	1.771
Presence of BCG scar	0.174	1.570	0.819	3.007
Taken BCG vaccine	0.186	1.758	0.762	4.055

TB: Tuberculosis; OR: Odds ratio; CI: Confidence interval

Table 4a

Association between sociodemographic characteristics and incidence of TB infection (n=345)

Variables	TST - Positive (%)	TST - Negative (%)	P
Gender
Male	24 (26.4)	67 (73.6)	0.038*
Female	43 (16.9)	211 (83.1)	(Chi-square value=4.26)
Present place of residence
Home stay	39 (29.3)	94 (70.7)	0.001*
Hostel stay	28 (13.2)	184 (86.8)	(Chi-square value=13.5)
Illiterate	4 (44.4)	5 (56.4)
Educational			0.011*
Primary	5 (35.7)	9 (64.3)	(Fisher’s exact test value=10.44)
Status
Secondary	19 (26.4)	53 (73.6)
Graduate	39 (15.6)	211 (84.4)
Married	27 (29)	66 (71)	0.020*
Married	27 (29)	66 (71)	0.020*
Marital status
Unmarried	40 (16)	210 (84)	(Fisher’s exact test value=7.226)
Separated	0	2 (200)

TB: Tuberculosis; TST: Tuberculin skin test

Table 5

Multivariate analysis for incidence of TB infection among the study subjects (n=345)

Variables	Multivariate analysis (P)	OR (odds ratio)	95% CI for OR
			Lower	Upper
Sex	0.244	0.699	0.383	1.277
Education	0.363	0.711	0.341	1.483
Work commute time	0.403	1.193	0.789	1.803
Marital status	0.540	0.810	0.413	1.588
Duration in hospital	0.530	1.946	0.993	3.816
Designation	0.834	0.968	0.714	1.313

TB: Tuberculosis; OR: Odds ratio; CI: Confidence interval

DISCUSSION

Prevalence of latent TB infection among the study subjects

This study found out that the overall prevalence of LTBI was 20.1% by TST method, and induration ≥10 mm is considered as positive. In a study conducted in Vellore by Christopher DJ et al. among 436 HCWs, it was found that the prevalence of LTBI was 47.8% among HCWs, considering ≥10 mm induration as positive.[3] This is higher than our study. The subjects in the above study consisted of nursing students, whereas in our study we enrolled HCWs from different work profiles. As per the TB statistics for India, published in 2015, the numbers of sputum-positive TB cases are more in Tamil Nadu when compared with Karnataka.[4] Hence, it may be reasonable to expect that the exposure to patients with TB is more among HCWs in Tamilnadu state. All these factors may be responsible for higher LTBI among HCWs in Vellore. In another prospective study conducted in India by Anand et al. in Tamil Nadu among 206 HCWs with the objective of screening of HCWs for LTBI in a tertiary care hospital, it was found that 36.6% were infected with TB using a TST induration ≥10 mm as a cut-off point.[5] This study also had higher TST-positive subjects compared with our study.

The prevalence of LTBI was significantly associated with age, place of residence, education, work commute time, work profile, mode of transport, and BCG vaccine. The higher the age, the more the risk of LTBI. This was mainly attributed to greater duration of work exposure at the hospital. A nested case–control study conducted at a tertiary teaching hospital in India by Mathew et al. with the objective of assessing the risk factors for TB among HCWs found that HCWs with frequent patient contact and those with body mass index <19kg/m2 were at high risk of acquiring active TB. Nosocomial transmission of TB was pronounced in locations such as the medical wards and microbiology laboratories.[6] In this study, incidentally, a study subject working in the chest medicine outpatient department who was negative when baseline TST was done subsequently developed active PTB. Another study subject (a nurse from the Gastroenterology Department) who was TST-negative at baseline developed extra-pulmonary TB. Our study found that illiterate subjects and subjects who had studied up to primary level of education had more chance of TST positivity. A longitudinal study conducted in inner Mongolia Autonomous Region of China by Guangxue He et al. with the objective of measuring the prevalence and incidence of latent TB infection (LTBI) among village doctors (unqualified – have no more than high-school education and less than 1 year of medical training), it found that 19.5% of the 880 participating village doctors had a positive TST and 46.0% had a positive QuantiFERON-TB Gold In-Tube test (QFT-GIT) result. The factors associated with TST prevalence were presence of BCG scar and smoking. The risk factors associated with QFT-GIT prevalence included being male, education lesser than college level, and working for more than 25 years as a village doctor.[7] This indicates that even among “village doctors,” poor education has a role to play in the prevalence of TST positivity.

Incidence of latent TB infection

Our study found that incidence of LTBI was 19.4% by TST method, induration ≥1 0 mm being considered as positive. This is far higher than that reported in a systematic review by Baussano et al., the objective of whose study was to assess the annual risk of LTBI among HCWs and the incidence rate ratio for TB among HCWs. This systematic review found that TB incidence was 8.4% in high TB burden countries.[8] Underreporting of TB cases from high burden countries may be the reason for finding lower incidence in the systematic review. In this study, the incidence of LTBI was significantly associated with gender, place of residence, education, work commute time, marital status, and duration of employment and work profile. The greater the duration of work at the hospital, the more the chance of TB infection. Other factors such as ventilation and infection control practices may also play a role in infection rates among these HCWs. Male gender had more TST conversion – this is possibly because male subjects tend to travel more compared with females, and hence they are at higher risk of acquiring LTBI. Subjects residing at home had more TST conversion in comparison to those staying in institution-provided hostels. These subjects stay far from hospital and need to use public transport to reach hospital. This exposes them to crowds, thereby making them vulnerable to acquire LTBI. Study subjects who were illiterate and studied up to primary educational level had more TST conversion. Lack of knowledge about hospital-acquired infections and inadequate use of personal protective measurements would have contributed to the incident LTBI. There was a significant association between work profile and TST 2 positivity. There was a greater conversion among the study subjects working in the housekeeping, nursing, and laundry departments. Many of these employees lack knowledge regarding personal protective measures which might have contributed to higher incidence of LTBI. Lack of cough etiquette like covering mouth while coughing, sneezing, and yawning still persists among many infective patients, and these behaviors might have contributed to developing LTBI among the nurses. In the case of housekeeping staff and laundry workers, it can be postulated that exposure to fomites such as linen used in the wards may have resulted in the incidence to LTBI.

CONCLUSION

The prevalence of LTBI among HCWs was 20.1%, and it was significantly associated with age, place of residence, education, work commute time, and mode of transport. The housekeeping staff and laundry workers (those who come in contact with patients) had a greater risk of having TB infection. Subjects who were without the BCG scar had greater risk of having a positive TST reaction. Multivariate regression analysis revealed that no single variable was associated with the prevalence of LTBI. After 1 year of follow-up, the incidence of LTBI was 19.4% and it was significantly associated with male gender, those who reside at home, lower educational status, longer work commute time, and married status; nurses and housekeeping staff had higher incidence of LTBI. Multivariate regression analysis revealed that no single variable was associated with the incidence of LTBI.

Recommendations

TB infection control programs should be implemented in healthcare settings. Administrative controls should be undertaken to take care of the health of HCWs. This could involve implementing a variety of activities, with a key component being baseline and serial screening for LTBI of HCWs at risk for exposure to TB. Although the incidence of LTBI was seen mainly among subjects who were nurses and those working in laundry and housekeeping departments (who came into contact with patients and who are involved in sanitation work and handle fomites), health education sessions to all the HCWs about occupational health safety and provision of equipment/PPE are the need of the hour. Strengthening of the ongoing occupational safety training sessions in the institution is also advised. Administrative controls could also include written TB infection control plans to ensure prompt detection, separation from others, and treatment of persons who have suspected or confirmed TB disease. Proper follow-up of daily reported cases and appropriate disposal of sputum and infective material are necessary. As per Revised National TB Control Programme (RNTCP) guidelines, hospitals should have cough corners/counters, and a specific staff should be posted for cough screening, separation, fast-tracking, providing face mask, and counselling.[9] Separate inpatient facility for bacteriological positive TB patients and other airborne infectious patients is recommended. The hospital management should provide N95 masks to the hospital staff in high-risk zones. The hospital should display posters and signs to remind patients and staff of proper cough etiquette and respiratory hygiene. The hospital could undertake engineering measures such as constructing negative pressure rooms and renovation of wards with poor ventilation. At the national level, with adequate budget allocation, the RNTCP should implement basic TB infection control measures and enhance capacity for TB infection control – this in turn will reduce exposure of HCWs to TB. Medical schools and teaching hospitals, in particular, have to adopt and translate the Tuberculosis Infection Control policy into concrete action.

Limitation

People infected with Mtb may have a negative reaction to the TST if many years have passed since they became infected. They may have a positive reaction to a subsequent TST because the initial test stimulates their ability to react to the test. This is commonly referred to as the “booster phenomenon” and may incorrectly be interpreted as a skin test conversion.[10] In this study, a two-step method was adopted and since follow-up was done after 1 year, first TST effect may not influence repeated TST reaction. In this study, the random sampling of doctors was not done as many of them refused TST. Required numbers of pharmacy workers could not be followed up completely as they refused repeat TST. False-positive results, even though rare, due to nontuberculous mycobacteria infection, may have occurred in some study subjects.

Financial support and sponsorship

This research was supported by Arkray Healthcare Pvt. Ltd (AMI), Gujarat, who provided 2 TU PPD.

Conflicts of interest

The Associate Editor, who is a co-author of this paper confirms that all due processes have been followed in the publication of this article.

Table 4b

Association between occupation characteristics and incidence of TB infection (n=345)

Variables	TST - Positive (%)	TST - Negative (%)	P
Work commute Time
≤10 min	24 (12.4)	169 (87.6)	0.001*
11-30 min	20 (32.8)	41 (67.2)	(Chi-square value=14.9)
>30 Min	23 (25.3)	68 (74.70
≤10 Years	51 (16.7)	254 (83.3)	0.002*
Duration of Employment hospital
11-20 years	12 (40)	18 (60)	(Fisher’s exact test value is11.897) (n=341)
>20 years	3 (50)	3 (50)
Doctors	12 (17.4)	57 (82.6)
Nurses	19 (12.3)	135 (87.7)
Laboratory technicians	8 (29.6)	19 (70.4)	0.009*
Work profile
Nursing aides	14 (25.9)	40 (74.1)	(Fisher’s exact test value=17.427)
Pharmacy	2 (22.2)	7 (77.8)
Laundry	4 (50)	4 (50)
Dietary	1 (16.7)	5 (83.3)
Housekeeping	7 (38.9)	11 (61.1)

TB: Tuberculosis; TST: Tuberculin skin test