Long hospital stays for TB treatment can increase risk of reinfection with MDR or XDR-TB strains

Hospitalising people with tuberculosis – whether it is drug-sensitive or drug-resistant - for long periods in order to treat the condition puts them at increased risk of reinfection with multidrug-resistant (MDR) and extensively-drug-resistant (XDR)-TB strains, according to a Latvian study presented at the World Lung Health Conference in Paris last week.

“To decrease risk of TB transmission in hospitals, strong infection control measures and ambulatory treatment treatment should be enforced,” said Anda Nodieva of the State Agency of Tuberculosis and Lung Disease of Latvia.

Background on MDR-TB in Latvia

Rates of TB, including MDR-TB, jumped dramatically in Latvia, along with other countries in the former Soviet Union, following the collapse of its centralised public-health system. The case rate peaked around 1998, with about 90 cases per 100 000 individuals. However, the burden of disease has been particularly heavy among the more vulnerable members of society (prisoners, people who abuse alcohol, the homeless etc).

However, in the mid-1990s Latvia began to implement a Directly Observed Therapy Short Course (DOTS) programme and cure rates began to improve. As of 2007, the case rate has fallen to 54.6 per 100,000 population.

MDR-TB continued to make up an increasing proportion of TB cases however, until a few years ago, when Latvia began to implement a DOTS-Plus programme (which includes second-line treatment with individualised regimens and specialised services required to deliver it).

The country has been working in partnership with the Green Light Committee, an international technical assistance initiative that has been set up to help facilitate access to second-line medications once a programme demonstrates that it can use them responsibly (in line with the WHO guidelines, with adequate diagnostic facilities and with patient support measures to decrease the risk of XDR-TB).

Latvia has since implemented a model MDR-TB control programme and the total number of MDR-TB cases was cut in half (see this report), with the overall percentage of MDR-TB down from 17.7% of all cases in 2004 to 10.1% in 2007. In fact, Latvia’s MDR-TB programme is often where health officials go to learn how to set up their DR-TB projects (such as the teams from Tugela Ferry, the site of the XDR-TB outbreak in KwaZulu Natal, South Africa).

TB drug resistance can either be acquired (due to poor adherence to treatment) or primary (when someone is infected with a strain which is already drug-resistant). The programme observed that the most significant decrease in MDR-TB was among those with acquired resistance, which is a good sign that their DOTS programme is working. Primary MDR-TB cases have been dropping as well but not as dramatically. In addition, researchers noted that a number of patients with a previous episode of TB were coming in for retreatment.

Analysing the recurrent TB cases

These cases could be due to relapse, failure or treatment defaults, so Nodieva and colleagues conducted a retrospective case control study in 54 retreatment cases from 2001 to 2007 where cultures were available for previous and current episodes. Specimens from these cultures were put through a battery of drug sensitivity and genotyping and spoligotyping tests.

Eleven of the cases were infected with identical TB strains in both episodes, suggesting that the cases were due to reactivation of the earlier case. Forty-three of the retreatment cases were infected with genetically different strains suggesting that they had been reinfected with a different strain.

So the researchers investigated the characteristics of the cases to see if there were any risk factors for reinfection compared to the reactivation controls.

Reinfection cases were more likely to have drug-sensitive TB in 30 of 43 cases(70%, OR 6.15, p<0.05) in their previous episode of TB, while the reactivation cases were more likely to have had multidrug resistant TB during their last episode in five out of 11 cases (45%, OR 0.19, p<0.05).

Of note, the median time to sputum smear conversion (an indication that TB is no longer infectious) for the first episode of TB was 80 days (for non-MDR) and 326 days for the one MDR case in the subset of 31 smear-positive reinfection cases, while it was between 127 to 141 days among the reactivation cases.

Despite the quick time to sputum smear conversion, the reinfection cases had spent a much longer time in inpatient care, (for instance, 229 days for 30 cases with smear-positive non-MDR-TB, while reactivation cases with smear-positive non-MDR-TB had stayed in hospital for only about 13 days after their sputum conversion). Seven smear-negative reinfection cases with non-MDR-TB stayed in hospital for 272 days.

Smear positive MDR-TB cases tended to stay in hospital longer — but this was perhaps part of the problem, because they may have the source of some of the reinfection cases. In fact, among the second episodes of the 43 reinfection cases, 36 were either MDR or XDR-TB. Clustering of molecular types was also common, which indicates recent transmission.

“So there appears to be nosocomial transmission of MDR-TB,” said Nodieva.

Towards better infection control and more ambulatory treatment

Many of the reinfected patients had spent their first episode in the hospital at a time when all the TB patients were mixed together (prior to 2004), Since that time, the clinic has tried to introduce better infection control practices. Several separate wards have been set up for confirmed XDR-TB cases, MDR-TB cases and possible cases (including contacts of MDR-TB cases), and for patients without a known risk of MDR-TB, and inside these wards sputum smear-positive patients are isolated. (The safest but far more expensive solution to reduce nosocomial transmission would be an individual room for each smear-positive patient).

As in much of the developed world, it is still policy to provide treatment to DR-TB patients in hospital while they are smear-positive “but now, just after sputum conversion, we discharge people for ambulatory treatment,” said Nodieva. Some patients with co-morbidities may need to remain in hospital longer or doctors may continue to keep patients in hospital if they live too far from the clinic to continue treatment or if they are homeless.

She told aidsmap.com that in some settings it would probably be better if all treatment could be given by DOTS in the person’s own home, but that in Latvia the socio-economic situation of many patients with MDR-TB limits their ability to do that: “They are on the street or somewhere,” she said, and in the hospital at least they have a roof over their head.

Note that this is a very different situation from southern Africa, where housing may be inadequate, but where people usually have homes — while it may be much more difficult to find a bed in a DR-TB hospital. And yet, when people with DR-TB are put into a facility, they may be kept there for a much longer period of time.

“But we should remember, that if we shorten the possible time that they could transmit in the hospital it is better,” Nodieva said — both to prevent nosocomial transmission and reduce the risk to healthcare workers.