Health authorities' failure to test tuberculosis (TB) patients for drug susceptibility appears to have inadvertently fuelled the spread of deadly drug-resistant strains of the disease in KwaZulu- Natal, scientists report in the Clinical Infectious Diseases journal.

Their research puts pressure on the health department to improve its monitoring of TB drug resistance and consider testing more patients before treating them.

Health authorities test only a handful of patients for drug resistance as testing is expensive and not widely available; guidelines say the tests should be administered only to patients who fail to respond to standard (first-line) treatment, or have been exposed to people infected with drug-resistant strains.

KwaZulu-Natal shot into the headlines last year when scientists announced that 53 patients in Tugela Ferry had succumbed to extremely drug resistant (XDR) TB, all but one of whom died. XDR-TB is resistant to almost all known drugs, and growing numbers of cases have since been detected in all nine provinces.

In KwaZulu-Natal for instance, 135 new cases of XDR-TB were identified last year, and another 105 cases in the six months to June.

Scientists have now tracked the evolution of drug resistance in the province between 1994 and 2002 by analysing samples of a highly transmissible strain called F15/LAM4/KZN circulating in the region.

They found that some patients were already resistant to treatment in 1995 when the World Health Organisation advised SA to implement a four-drug combination for treating first-time TB patients -- isoniazid, rifampicin, pyrazinamide and ethambutol.

This meant that some patients were inadvertently getting a drug cocktail with only one or two effective ingredients, and so they rapidly developed resistance to these drugs too.

"I said to the health department then that it was a recipe for disaster," said study co-author Willem Sturm, of the University of KwaZulu Natal's Nelson Mandela School of Medicine.

"No one listened, and multi-drug resistance (MDR) strains spread." He said that drug resistant strains were more easily spread between people than scientists initially thought.

Six years later, the same mistake was made when SA introduced World Health Organisation-approved guidelines stipulating which medicines to use for treating MDR-TB, and the absence of routine susceptibility testing for second-line drugs this time led to the emergence of XDR-TB.

Patients got either ethambutol or cycloserine, in combination with pyrazinamide, ethionamide, kanamycin, and ciprofloxacin or ofloxacin. Unknown to health authorities, there was already resistance to both ethionamide and ethambutol, dating back in fact to 1997. The scientists traced kanamycin resistance back to 1999, and fluroquinolone resistance back to 2000.

The researchers also found resistance to streptomycin as far back as 1994. The drug is added to the standard four in first line treatment for patients who have interrupted and then recommenced treatment.

The health department's TB head, Dr Lindiwe Mvusi, said routine susceptibility testing would be impractical. "It's just not going to be possible to test everyone," she said.

Mvusi said the health department's most recent drug resistance survey was conducted in 2001. A new study was to commence at the end of this year.