Married to M.tuberculosis: risk of infection and disease in spouses of smear positive tuberculosis patients

Abstract

Objectives

To quantify the risk of infection and disease in spouses of tuberculosis patients, and the extent to which intervention could reduce the risk in this highly exposed group.

Methods

We compared HIV prevalence, TB prevalence and incidence and tuberculin (TST) results in spouses of TB patients and community controls. HIV positive spouses were offered isoniazid preventive therapy(IPT), and TST was repeated at 6, 12, and 24 months.

Results

Of 399 smear positive patients ascertained prospectively, 201 reported cohabiting spouses, with 148 recruited. Five (3%) had active TB. We identified 203 spouses of 406 previously diagnosed smear positive patients. 11 had already had TB, and the rate of TB was 2.4 per 100 person years(py) over two years (95% ci 1.15-5.09). 116 were found alive and recruited. HIV prevalence was 37% and 39% in the prospective and retrospective spouse groups and 17% in controls. TST was >=10mm in 80% HIV negative and 57% HIV-positive spouses ascertained retrospectively; 74% HIV negative and 62% HIV positive spouses ascertained prospectively, and 48% HIV-negative and 26 % HIV-positive community controls. 18/54 HIV positive spouses completed 6 months IPT. At two year follow-up, 87% of surviving spouses had TST >=10mm and the rate of TB was 1.1 per 100py (95% ci 0.34-3.29).

Conclusions

Spouses are a high risk group who should be screened for HIV and active TB. TST prevalence was already high by the time the spouses were approached but further infections were seen to occur. Uptake and adherence to IPT was disappointing, lessening the impact of short duration therapy.

Introduction

Although most Mycobacterium tuberculosis (M.tb) infection in endemic areas is acquired from sources unknown to the individual (Crampin et al., 2008, Crampin et al., 2006, Wilkinson et al., 1997), close contact with a smear positive tuberculosis (TB) patient greatly increases the risk of infection and subsequent disease (Crampin et al., 2004, Sinfield et al., 2006). During contact investigations in rural Malawi, 5% of adult TB cases reported a spouse with prior TB (Glynn, 2007). Wives are particularly likely to have intense exposure to M.tuberculosis, through daily close contact and nursing. The spouse may also be exposed to other TB patients, if they accompany their partner to outpatient facilities or provide nursing care in the hospital ward. In this setting HIV prevalence in TB cases is 60% or higher (Crampin et al., 2009) so, if around two thirds of spouses have concordant HIV-positive status, HIV prevalence among the spouses would be expected to be around 40%. Spouses of smear positive patients are thus a high risk group, and they are also an easily identifiable group who can be accessed through their nursing role for assessment of risk of TB and offered preventive therapy. The spouse-spouse tuberculosis association and the risk of prevalent tuberculosis and HIV amongst spouses of TB patients have been described (Jurj et al., 2006, Suggaravetsiri et al., 2003) but the preventable risk in subsequent years has not been quantified.

We identified spouses of newly diagnosed smear positive pulmonary TB patients for investigation of infection and disease, with active follow-up and preventive therapy for those who were HIV-positive. We also identified a cohort of spouses retrospectively from patients diagnosed in the previous four years to allow calculation of rates of TB in the absence of any specific intervention. We describe the risk of infection and disease and associated risk factors in spouses and in a control group selected from the community.

Methods

The study was conducted as part of the Karonga Prevention Study (KPS) in northern Malawi. Staff have seen all TB patients since 1986. Sputum smears and cultures are processed in the project laboratory. For this study all district residents with smear positive pulmonary TB were identified.

Prospective recruitment

From November 2002 to July 2005, smear positive TB patients were questioned on marital status and co-residence with any spouse since onset of cough. Spouses acting as “guardians” (nurses) on the TB ward were contacted directly. When the spouse was absent, patient consent was sought to visit the household to recruit the spouse.

Retrospective recruitment

Smear positive TB patients diagnosed in the four year period prior to the recruitment phase (January 1999 to October 2002) were identified from project records and traced by KPS field staff members known to them (household visits to TB patients had previously been routinely conducted). Surviving TB patients or appropriate informants were asked about the patient’s marital status and co-residence of any spouses at the time of the index patient’s illness.

Control recruitment

Community controls were selected thoughout district using a random frequency matched field sampling method as previously described (Crampin et al., 2001). This gives a random sample with an age, sex and area distribution similar to that of TB patients.

Baseline visit

Spouses identified prospectively were visited 6 weeks after the TB patient had started treatment. Spouses identified retrospectively were visited during the first year of the study, providing a wide range of follow-up times. Baseline visit procedures were identical.

KPS field staff (medical assistant or trained health assistant) traced the spouse to their current household, checked eligibility (including cohabition at the time of the index case illness) and collected outcome details for those that had died, left the district, or been treated for TB in the intervening period. Written consent was obtained for interview, clinical examination, CD4 count, tuberculin skin test (TST), HIV counselling and testing, syphilis testing, pregnancy testing, and follow-up where appropriate. The TST was conducted using the Mantoux method with 2 I.U. intradermal RT23. Spouses were invited to attend the hospital for clinical review by a medical officer and chest radiography to exclude active TB. Those with productive cough for three or more weeks had sputum collected.

At the time of post-test counselling for HIV, those who were HIV-positive with induration size less than 5mm had a second skin test with Candida antigen to determine ability to mount a delayed type hypersensitivity response, and HIV-positive participants were offered 6 months of isoniazid preventive therapy (IPT) irrespective of TST result, given their recent exposure.

Controls were selected during the recruitment phase, and written consent obtained for interview, clinical examination, venepuncture and TST. Procedures were identical to those for spouses except that CD4 counts were only performed on a sample.

Antiretroviral therapy (ART) was not introduced to the district until June 2005; at that time, any HIV-positive individuals in the study eligible on the basis of their clinical condition (WHO Stage III/IV) or CD4 count (< 250/ul ) were revisited and referred to the ART clinic.

Follow-up visits

Prospectively recruited spouses were followed up at 6, 12 and 24 months from baseline date. On each occasion the same procedures were repeated. Unless there was a specific indication, they were not referred for clinical review. If at any point (or if notification was made between visits) the spouse was found to have died, left the district, withdrawn consent, been exposed to TB again by a spouse or developed TB, outcomes were recorded.

Retrospectively recruited spouses and controls were only followed up if they had been started on IPT for monitoring purposes; these data are not presented. Additional cases of TB were passively ascertained from routine KPS activities after the 24 month follow-up was complete, to December 2009.

Re-exposure of spouses followed prospectively

If a spouse was identified as having been re-exposed to a spouse with smear positive TB during follow-up (because the index patient had recurrent TB or because a new spouse or concurrent wife had TB) the spouse was considered to be censored from the study as the risk of the first contact could not be assessed.

Isoniazid preventive therapy

IPT was offered to all HIV-positive spouses recruited prospectively irrespective of TST result. IPT was only offered to HIV-positive spouses recruited retrospectively or to controls if they had a TST induration >=5 mm, as there was no known recent exposure to TB. Those on IPT were followed up monthly at home for the 6 months of treatment. They were interviewed and examined on each occasion, and pill counts performed, and the next month’s supply was provided.

Data handling and analysis

Data were double entered, verified and amended as appropriate. TST “positivity” was interpreted as a proxy for likely infection with M.tuberculosis with two cut-offs; induration >=10mm, and >=15mm (Centres, 2000). TST conversion was defined as a change from <10mm to >10mm with an increase of 6mm or more (Society, 2003). The same TST cut-offs were used for HIV-negative and HIV-positive individuals as our data supported findings elsewhere that reduced cut-off values for HIV-positive individuals are of limited benefit (Cobelens et al., 2006).

Active disease (laboratory confirmed or clinically diagnosed) was defined as either sputum (or other biological specimen) smear or culture positive; symptoms of TB combined with chest x-ray appearance consistent with TB; or clinical symptoms of TB not responding to antibiotic therapy which improved on TB treatment. Onset of exposure was defined as the date of collection of the first sputum positive on microscopy or culture from the index case. In the survival analysis, data were censored at time of loss to follow-up, departure, death, or end of the study; and, for prospective spouses, if re-exposed to TB in a spouse.

Ethical approval

Ethical approval was granted by the National Health Sciences Research Committee of Malawi and the London School of Hygiene and Tropical Medicine Ethics Committee.

Results

Spouses seen prospectively

During the recruitment phase of the study, 399 individuals were identified with smear positive TB, (Figure 1a). These individuals reported 201 cohabiting spouses of whom 183 could be traced. Of those traced, three had died, 4 had left the district, 9 refused to take part in the study and 15 were reported to be alive but were not found. A further 4 denied close contact with the patient at the time of their illness, leaving 148 spouses with a history of TB exposure who are considered in further analyses.

Figure 1.

Prospectively recruited spouses

103/148 spouses were recruited within one month of the index patient’s first smear or culture positive samples, and 34 between one and six months. For the remaining 11, exposure was known to have started up to 18 months before recruitment due to treatment being started very late (as a result of smear negative culture positive patients becoming smear positive later, or patients defaulting prior to treatment). Sixteen (11% of cases) had been coughing for more than three months before their first positive sample.

99 (67%) of the spouses were female, and the women were younger than the men (median ages 34 and 39 respectively). 74/134 (55%) of the TB index cases with identified spouses were HIV-positive (14 were not tested). Three spouses refused HIV testing. Of the remainder, 54/145 (37%) were initially HIV-positive, and 4 seroconverted during the course of the study, despite counselling.

Of the 54 initially HIV-positive spouses, 50 had CD4 counts available at baseline. Of these, 15 (30%) had a CD4 count <250, (eligible for ART under subsequent Malawi programme criteria), and 26 (48%) had CD4 counts <350.

By definition, those included had lived in the same household as the index case when they were sick, but 98% had also been sleeping in the same room and 20% had nursed the index case. Wives were twice as likely as husbands to have taken a nursing role.

Follow-up

Of the 148 individuals who had a first baseline visit, 119 had at least one follow-up visit. By the end of the study period (24 months after recruitment), 5 spouses had died, 12 had left the district, and 6 were re-exposed to TB. Thirty-eight spouses withdrew during the study and 87 remained enrolled for 24 months.

Of the 54 HIV-positive spouses in the prospective study, 5 were not offered IPT as they were symptomatic (and were subsequently treated for TB), and 15 refused, died before starting treatment or had contraindications. Thirty four started on IPT, though only 18 completed the six month course.

Spouses seen retrospectively

406 individuals were identified as having had one or more episode of sputum smear positive pulmonary tuberculosis from January 1999 to October 2002 (Figure 1b). For 196 of them, one or more spouse who had been living with the individual at the time of the TB could be identified, giving 203 spouses to be considered in further analyses.

Figure 1b.

Recruitment of retrospective spouses.

Before the recruitment period, 11 (5 %) had already developed TB (6 HIV-positive, 5 unknown). 22 (11%) had died (including 4 of those who had TB). 29 (14%) had left the district, and 3 (1%) were untraceable or not included for other reasons. Of the 142 who were found, 18 (13%) refused and 124 (87%, 61% of the overall sample) were recruited. On interview 116 of these went on to be included in the study. Eight were excluded as they reported that they had not cohabited with the TB cases at the time of illness or for other reasons.

Of the 116 surviving spouses included in the study, 56% were female. Females were younger than males (median 34 and 43 years). 93% agreed to HIV testing and 42/109 (39%) were HIV-positive. CD4 counts were available on 39; 11 (28%) were < 250, 1 (3%) was 250-350 and 27 (69%) were >350. 14/116 (12%) of the spouses had nursed the case and 113 (97%) had shared a sleeping room.

Controls

811 controls were initially recruited and 8 were excluded from this analysis as they reported household contact with a spouse with TB in the 5 years prior to interview, leaving 803 to be considered in the analysis. 53% were female, and females were younger than males (median 32 and 39). 71% agreed to TST, and 91% had an HIV-test, of whom 17% of were HIV-positive.

Risk of infection and duration of exposure

Prospective

70/91 HIV-negative and 47/54 HIV-positive prospectively recruited spouses had a TST available at first visit and the majority were “positive” at a cut-off of 10mm (Table 1). Median induration size was 13 mm in HIV-negative spouses and 12 mm in HIV-positive spouses, or 15 mm for both HIV-positive and HIV-negative if zero indurations were excluded. 10/17 of those who were HIV-positive and had an RT23 induration size less than 5mm agreed to a further skin test with Candidin antigen. All had some response to Candidin and 8 had indurations >=10mm. Individuals with low CD4 counts were more likely to have a zero TST induration (data not presented) but there was no difference in the distribution of the responses above zero, by CD4 count, consistent with TST response being an “all-or-nothing” phenomenon.

Table 1.

TST at baseline groups of individuals

	TST >=10mm		TST >= 15mm
	n/N	%	n/N	%
Spouses seen prospectively
HIV+	29/47	62	19/47	40
HIV −	52/70	74	29/70	41
Spouses seen retrospectively
HIV+	21/37	57	11/37	30
HIV−	50/60	80	24/60	40
Controls
HIV+	25/96	26	13/96	14
HIV−	219/456	48	85/456	19

Despite initially having a lower, prevalence of TST “positivity” than men, “positivity” in female HIV-negative spouses increased with increasing time since the index case was known to be smear or culture positive (Table 2). This reflects the variation in duration of symptoms prior to diagnosis but also includes follow-up sampling so captures ongoing TST conversions after the index patient had started treatment.

Table 2.

Proportion TST positive (>=10mm) by duration of exposure prior to TST of spouses in prospective study by HIV status.

		<0-2 months		3-4 months		5-9 months		9-12 months		P value of chi square for trend
Spouse HIV+	Male	5/5	50%	4/8	50%	4/7	57%	3/7	43%	0.9
	Female	15/25	60%	10/15	67%	10/11	91%	5/8	63%	0.3
Spouse HIV −	Male	18/20	90%	15/17	88%	16/18	89%	17/19	90%	1.0
	Female	29/44	66%	18/25	72%	33/36	92%	23/23	100%	< 0.001

Seven HIV-negative spouses and six HIV-positive spouses had TST conversion during the course of the study. All six of the HIV-positive spouses who converted were on IPT.

Retrospective

60/67 of HIV-negative spouses and 37/42 HIV-positive spouses agreed to TST (table 1), with median induration of 14 mm and 12 mm respectively (14 mm in both groups when restricted to those with more than zero induration). The interval between exposure to the smear-positive TB case and the spouse being seen ranged from 9 to 58 months, median 33 months. Surviving HIV-negative spouses from the retrospective cohort, also had TST prevalence of 80% (Table 1) confirming the extremely high risk of infection among intimate contacts.

Controls

The median TST induration was 9 mm for HIV-negative controls and 0 mm for HIV-positive controls (13 mm and 14 mm respectively when restricted to those with any response). Prevalence of TST “positivity” was considerably lower than in the exposed spouses (Table 1).

Rate of tuberculosis

Prospective

On clinical screening, 12/148 (8%) spouses reported cough of more than three weeks duration, and sputum was collected from four with productive cough, three of whom were started on TB treatment immediately. One was smear positive, the other two were later found to be culture positive. The fourth (who had TST of 16mm and was also culture positive) defaulted but presented for TB treatment two years later and died soon after. One of the culture positive individuals was HIV-negative, the other three were HIV-positive. An additional HIV-negative individual had cervical lymphadenopathy consistent with glandular TB, and was started on treatment after culture confirmation from a fine needle aspirate. Three of the five TB cases were in men. These cases were used in calculation of disease rates since exposure.

No-one developed active TB during the two year follow-up period, but a further seven individuals (four women, three men) have been diagnosed with TB since this time: 6 were HIV-positive (two had smear positive pulmonary TB, two smear negative culture positive pulmonary TB, two clinically diagnosed pulmonary TB and one clinically diagnosed extra pulmonary TB). Four of the HIV positive patients had TST at baseline of >10mm, one had 0mm and the other refused TST.

Five of the six HIV-positive individuals who were diagnosed after the end of the study period, had taken IPT during the study, 3 had a full 6 months course, 2 had stopped after 2 months due to side-effects. The remaining person refused chest x-ray at baseline so active TB could not be excluded and was therefore not offered IPT.

Retrospective

Rate of disease was calculated using the 11 (5 %) who had already developed TB by the time of baseline visit (with survival times for those not found given by informants). Since the study visit, an additional 11 are known to have developed TB. Nine were HIV-positive and 2 were unknown HIV status, Three of the HIV-positives spouses had been offered IPT at the time of the recruitment into the study as they had a TST >5mm, however 1 only took the IPT for 2 months and the others refused.

Comparison of tuberculosis rates in prospective and retrospective spouses

Rates were calculated overall as HIV status of many individuals in the retrospective cohort were unknown. Rates of TB (laboratory confirmed and clinically diagnosed) were highest in the first six months after onset of exposure, (4.54 and 2.81 per 100 person years respectively for retrospective and prospective spouses), Table 3. In the intervention group the rate was extremely low for the period of observation although a cluster of cases were diagnosed after the end of the observation period. There was a statistically significant difference between the rates of disease between the prospective and the retrospectively recruited spouses in the first two years after exposure (log rank test for equality of survivor functions, p=0.0009). The cumulative hazard of TB in the two groups is presented graphically in Figure 2.

Table 3.

Rates of TB per 100 person years (95% ci) after exposure, comparing spouses recruited retrospectively and those with active screening and preventive therapy.

	Overall in first two years	<6 month	6 months to two years	Two to four years
Retrospective	2.43 (1.15-5.09)	4.54 (1.73-12.3)	1.49 (0.48-4.61)
Cases	7	4	3	5
HIV+	4	2	2	3
HIV−	0	0	0	0
HIV?	3	2 (both died	1 (died)	2 (1 died)
Prospective	1.06 (0.34-3.29)	2.81 (0.71-11.1)	0.47 (0.07-3.35)
Cases	3	2	1	8
HIV+	2	1 (died)	1	7 (3 died)
HIV−	1	1	0	1

The period of follow up extends beyond the end of the 24 follow up in some cases, as there was a long exposure prior to recruitment.

Figure 2.

Cumulative hazard of TB after onset of exposure.

Discussion

In this study we describe an extremely high risk group – spouses of smear positive pulmonary TB patients. They have a high prevalence of HIV, many are already immunocompromised and they have intense contact with infectious TB, sharing sleeping rooms and frequently also nursing the patient. This intense exposure and susceptibility is highlighted by the high rate of initial TST positivity at baseline compared to a control group, the subsequent additional TST conversions, the prevalence of active TB on screening and subsequent risk of TB. The TST cut-off of >=10mm, although widely used, will include more false positives, particularly in this area with ubiquitous exposure to environmental mycobacteria and high rates of BCG vaccination (Crampin et al., 2009). With this sample of adults, however, presence of a BCG scar (approximately 80% of participants) was not associated with significantly larger TST indurations. The higher TST cut-off gives a minimum estimate of M.tuberculosis infection and the “real” prevalence will lie between the two.

Ongoing TST conversions after the index case had started treatment, may be due to delay in the development of the hypersensitivity but may also be due to ongoing or increased exposure as the spouse accompanies the TB patient to outpatient cough clinics and nurses them as an inpatient on an open TB ward. This increasing exposure around the time of diagnosis is supported by the observation that the increase in rates and the TST conversions were more common in female than male spouses who are more likely to have been involved in care. Additional risk from health facility exposure is also consistent with other work in the same setting where 5 of 13 spouse pairs with active TB had different M.tb strains (Crampin et al., 2006). , Although women seemed more likely to become infected around the time of exposure, they did not seem to be more at risk of developing disease despite similar proportions of men and women receiving IPT.

The vulnerability of this group to co-infection with HIV and M.tuberculosis and the high risk of disease suggest that even in resource poor settings where household contact tracing is not feasible, spouses form a high risk and easily identifiable group for HIV counselling and testing and screening for active TB. Early identification of co-infected individuals is key to tackling the co-epidemic (Harries et al., 2010). In this group, later TB occurred despite a period of IPT, reaffirming the view that, if given, IPT should be lifelong for HIV-positive individuals. Despite the careful observation of our intervention group which is likely to result in improved ascertainment, we still identified lower rates of TB than in the historical comparison group. This suggests (given the limitations of our small sample size and some loss to follow up in both groups) that we averted a proportion of cases, although as HIV status was not known for many of the retrospective spouses who were not available for recruitment, the groups may not be directly comparable. In contrast to the good compliance with IPT in spouses from a different setting (Ngamvithayapong et al., 1997), we highlight compliance problems in this “well” group despite their intimacy with TB cases. Screening for active TB in spouses at the time of index case diagnosis was well received, but intense, repeated counselling akin to that offered with ART may be required to support extended IPT.