A South African doctor who witnessed the world's biggest XDR TB outbreak says ill-equipped hospitals means many patients are better off at home.
For a hospital in the poor, mountainous uMzinyathi district of rural KwaZulu-Natal, South Africa, the Church of Scotland Hospital in Tugela Ferry is impressive.
Housed in a modern building, its administration centre is equipped with the latest office equipment and computers linked to an intranet.
Nurses move across the well polished floors of the wards in their neatly ironed uniforms and in the grounds, patients take the opportunity to bask in the sun.
Kwanele Mpungose is the only patient in the 10-bed male isolation ward where he is being treated for multidrug-resistant (MDR) tuberculosis. His legs hurt too much for him to sit up, and the food on his bedside table remains untouched.
Mpungose, who is 32, has been in and out of hospital for more than a year. "I am happy here, I feel I am cared for," he says. "At home, I feel lonely as there is no food and there is no one to cook for me when I take my pills. My family is 25km away but they do not come to see me," he adds.
Despite being frail, Mpungose is determined to talk and seems grateful for the company. In another isolation ward, four HIV-positive women, also with MDR TB are sleeping peacefully. They are too weak to talk.
The isolation wards have been set up to monitor patients who cannot be admitted to the provincial MDR TB specialist centre Greytown Hospital or to King George V hospital 200km away.
They did not exist in 2006 when the Church of Scotland hospital recorded the world's largest outbreak of drug-resistant TB. In total, 266 people were diagnosed with extensively drug resistant (XDR) TB, the most lethal strain of tuberculosis, which does not respond to most drugs and can take up to two years to treat, and 205 cases of MDR TB, which responds poorly to standard treatment.
Nearly all the patients with XDR TB died, as well as eight members of staff. Many of the dead had been HIV-positive patients whose immune systems could not cope with the infection. If a patient with HIV contracts an XDR strain, the prognosis is particularly poor.
Hospitals ideal for TB spread
Dr Tony Moll, Chief Medical Officer at the Church of Scotland hospital is one of the doctors researching the outbreak. He has been working here for more than 20 years, and despite efforts to separate those with HIV from those with TB, he says it is hard to prevent patients picking up TB in hospital.
"At the bedside, I cannot tell the difference between MDR, XDR and normal TB," he explains. "You have HIV patients within surgical wards, HIV-positive mothers delivering babies in the maternity ward, it is difficult to separate them." By the time TB tests have been done, that patient might already have passed on the infection.
What's more, many hospitals provide an ideal environment for TB to spread. Research published in the Lancet medical journal outlines the dangers of poorly-ventilated, mixed general wards.
Using models, based on findings at the Church of Scotland hospital, it suggests nearly half of the predicted 1,300 new cases of XDR TB in the Tugela Ferry region by 2012 will be transmitted in hospital. The key to preventing this, it says, is a combination of comprehensive infection control measures and treating more TB patients in the community.
Since the deadly 2006 outbreak, the Church of Scotland Hospital has invested in measures to cut airborne infection. In addition to setting up isolation wards, it has hired an infection control policy officer who is tasked with monitoring infection controls.
Staff must wear masks when working in wards, windows are open at all times, and mechanical ventilation sucks air in and out of the wards. Signs dotted around the hospital urge staff and patients to comply with infection control measures.
Even so, it is not ideal. "We don't have the ideal situation to keep a person in hospital for the full duration of their treatment, we don't have the ideal circumstances to keep people in a patient-friendly environment", said Dr Moll.
He is piloting a community treatment programme in which nurses give daily injections to more than 87 patients at home. Patients like Kwanele Mpungose are supposed to be kept here only as they wait to be transferred to specialist MDR-TB treatment centres or when they are too weak to go home.
Lock up as a last resort
In many areas of South Africa, hospitals have responded to the public-health threat posed by drug-resistant TB by locking up patients in high-security, isolation hospitals, but Moll disapproves: "Detaining people promotes stigmatisation; a patient will say, 'I would rather die at home than die in detention behind a barbed wire, search lights and security guards'.
"We believe that looking after a patient at home will give us better treatment outcomes. The patient is in his own home, getting the same medicine that he would get in hospital. What we have established is that there is little community spread [of TB] and what we do know is that there is a lot of hospital spread. For staff and patients, it's much [more] dangerous to be in hospital."
Patients on the community treatment programme are encouraged to take simple measures to stop the spread of TB. This includes sleeping in a separate room from the rest of the family and speaking to them outside the house where there is plenty of fresh air.
This is not without its challenges because many people living in the rural areas around the hospital live in traditional huts with a small entrance, one small window and poor air circulation.
Moll admits there are those whose irresponsible behaviour puts the public at risk, such as a taxi driver with XDR TB who did not take his medication and was driving around with more than 10 people inside, very probably with the windows closed. Even so, he believes locking up patients should be used only as a last resort.
This feature draws on findings in the study "Prevention of nosocomial transmission of extensively drug-resistant tuberculosis in rural South African district hospitals: an epidemiological modelling study" published in The Lancet (Vol. 370, Issue 9597, 27 October 2007, Pages 1500-1507) by Sanjay Basu, Jason R Andrews, Eric M Poolman, Neel R Gandhi, N Sarita Shah, Anthony Moll, Prashini Moodley, Alison P Galvani, Gerald H Friedland.