Moderate Physician Experience Reduces TB Death Risk

TORONTO, Sept. 26 -- Experience in treating tuberculosis, even if it is outside a physician's main area of expertise, appears to play a key role in patient survival, according to researchers here.

In a retrospective observational analysis, 16.5% of patients managed by doctors with limited TB experience died within a year of diagnosis, compared with 6.2% of patients managed by physicians with greater experience, according to Kamran Khan, M.D., of St. Michael's Hospital here.

That mortality difference was statistically significant at P<0.001, Dr. Khan and colleagues reported in the Sept. 26 issue of the Canadian Medical Association Journal.

As TB incidence rates fall in the developed world, fewer physicians have hands-on experience with the disease, Dr. Khan and colleagues noted. In other areas of medicine -- such as HIV/AIDS or surgery -- physician experience has been shown to affect outcomes, but little had been known about the effect on TB.

To fill the gap, they analyzed all reported cases of TB in Toronto from July 1, 1999, to June 30, 2002 -- a total of 1,154. They were managed by 178 physicians, and the analysis included extensive clinical data on the patients as well as information on training and clinical experience of the treating physicians.

Doctors were considered to be TB specialists if they had graduate training in pulmonology or infectious disease and "treatment-experienced" if they saw on average at least one active case a year.

The study found that patients managed by TB-experienced doctors were more likely than those with less-experienced physicians (fewer than one patient per year):

• To get directly observed therapy: 46.8% versus 28.4%, which was significant at P<0.001.
• To have drug-resistant infections: 9.9% versus 3.9%, significant at P=0.02.
• To be treated by a physician trained in pulmonology or infectious diseases: 89.5% versus 29.2%, significant at P<0.001.

When the researchers considered all-cause mortality, they found that directly observed therapy and physician experience both had positive effects, but physician specialty did not. Specifically:

• Use of directly observed therapy reduced the risk of death by nearly 80%. The hazard ratio was 0.22, with a 95% confidence interval from 0.13 to 0.39, which was significant at P<0.001.
• Of note, 16.5% of the patients managed by TB-inexperienced physicians died within one year of diagnosis, compared with 6.2% of the patients managed by TB-experienced physicians (hazard ratio: 0.98, 95% confidence interval: 0.97-0.99, P< 0.001).
• When further stratified by the number of tuberculosis cases treated, 19.2% of the patients (19 of 99) managed by physicians with no previous tuberculosis experience during the previous three years died within one year of diagnosis. This compared with the deaths of 7.1% of the patients (31 of 435) managed by physicians treating up to five cases per year, and 5.7% of those (35 of 620) managed by physicians treating more than five cases per year.
• Being a TB specialist did not significantly affect the risk of death. The hazard ratio per additional case a year was 0.46, which was not significant.

The study "demonstrates that patient survival is influenced not only by patient characteristics, but also by features of the treating physician," Dr. Khan and colleagues reported, adding that two modifiable factors -- directly observed therapy and physician experience -- can improve outcomes for the disease.

In an accompanying commentary, Richard Long, M.D., of the University of Edmonton in Alberta, said "these conclusions are not particularly surprising" given recent demographic changes seen in TB prevalence.

But, Dr. Long said, the study offers an opportunity to fine-tune clinical practice in ways that will reduce all-cause mortality in TB patients -- namely by using directly observed therapy and referring patients to experienced TB doctors.

In the province of Ontario, where the study was conducted, there were 662 tuberculosis cases and 439 family practitioners in 2004, Dr. Long noted, so that a family doctor might meet a case of tuberculosis "as infrequently as every 15 to 20 years -- hardly enough to maintain vigilance."