Researchers found that pediatricians who read a fictional case report of a toddler with a leg fracture were more likely to suspect abuse if the child was described as coming from a lower-income family.

The hypothetical child's race, on the other hand, did not appear to influence doctors' opinions.

The second finding is somewhat surprising, according to the researchers. Studies looking at real-world cases have found that minority children are more likely to be evaluated for abuse than white children are.

And it's well known that the child welfare system in the U.S. has a disproportionate number of minority kids.

"There's very strong evidence of a racial difference in how patients are handled," said lead researcher Dr. Antoinette L. Laskey, a pediatrician at the Indiana University School of Medicine in Indianapolis.

But, she told Reuters Health, the reasons for that have not been clear -- including whether doctors may act based on unconscious racial stereotypes.

The current results suggest "there's more than race involved," Laskey said.

She was also quick to say, however, that the study doesn't mean pediatricians are consciously "classist" or otherwise biased when evaluating children's injuries.

The study, reported in the Journal of Pediatrics, included 2,100 U.S. pediatricians who responded to a survey that described one of four hypothetical cases.

All cases included an 18-month-old with an "ambiguous" leg fracture -- a type that can be caused by abuse or an accident.

But the cases varied by the child's race (black or white) and the family's economic situation; parents were described as having either professional jobs (accountant and bank manager) or working-class jobs (grocery clerk and factory worker).

Race had little effect on the doctors' responses. The study found that when the child was black, 45 percent of doctors believed there had "possibly" or "almost certainly" been abuse; another 32 percent were "unsure." If the child was white, 46 percent of pediatricians suspected abuse, with 28 percent saying they were unsure.

In contrast, there was evidence that parents' job descriptions swayed doctors' opinions.

When the child's family was lower-income, 48 percent of pediatricians thought there'd been abuse, versus 43 percent when the family was higher-income.

It's hard to know whether doctors' responses to a fictional case would be the same in real life.

And it's not clear, according to Laskey, whether attitudes about socioeconomic status might explain some of the racial differences in child abuse reporting seen in earlier studies.

She also stressed that she does not think pediatricians are consciously basing their diagnoses on parents' job titles. But in general, unconscious stereotypes can influence anyone's thinking.

"People tend to think that child abuse, or domestic violence, doesn't happen in upper-middle-class families, but of course it does," Laskey said.

It's important, she said, for doctors to be aware that unconscious generalizations could get in the way of diagnosing child abuse -- either missing it in kids from affluent families, or over-diagnosing it in children from poorer or minority families.

"My big take-home message for doctors is that we need to rely on the objective data," Laskey said.

It is true that studies have found children in poorer families to be at greater risk of abuse. But the poverty, itself, is not a "causative factor," Laskey said.

"Race and socioeconomic status shouldn't be things used in a diagnosis of abuse," she said.

SOURCE: bit.ly/wVlYrX Journal of Pediatrics, online January 5, 2012.

• A Permanent Economic Emergency