(Reuters) - Home visits and doctor's office interventions to prevent child abuse appear to have only limited success, with evidence mixed on whether they help at all, according to a U.S. analysis based on ten international studies.

As a result, the government-backed U.S. Preventive Services Task Force (USPSTF) said this week that current evidence is "insufficient" to recommend such programs for dealing with the hundreds of thousands of children reported to be abused each year.

"There have been a few studies done... (but) there's inconsistency in the results across these trials," said David Grossman, from Group Health Research institute in Seattle who is a member of the USPSTF panel. "I wish we could be more definitive on this."

According to the U.S. Department of Health & Human Services, about 675,000 children were reported as victims of child abuse or neglect in 2011, just under one percent of children nationwide. About nine percent of those were sexually abused.

In a review conducted for the USPSTF, researchers from Oregon Health & Science University in Portland analyzed 10 studies of child abuse prevention programs involving home visitors in the United States, UK and New Zealand, and one intervention done in a Baltimore clinic.

These studies included families at higher risk of child abuse based on responses to questionnaires or referrals from midwives and nurses.

For the one intervention that didn't involve regular home visits, at-risk families were given materials about child abuse and referred to social workers.

Some of the programs were linked to a drop in families being reported to authorities such as Child Protective Services and children's emergency room visits, according to findings published in the Annals of Internal Medicine, but others found no such link.

"It looks like some progress is being made in terms of evaluating early childhood abuse prevention programs," said Elizabeth Letourneau, who studies child sexual abuse at the Johns Hopkins Bloomberg School of Public Health.

There's a need for more of those evidence-backed programs, she said, but also to focus on sexual abuse in particular, an area which is especially lacking when it comes to research.

There's evidence that at least one program, focused on teen mothers, can help prevent child abuse, said Jane Barlow, a child abuse researcher at Warwick Medical School in Coventry, UK.

The best time to act to prevent abuse and neglect, Barlow told Reuters Health, seems to be while at-risk women are pregnant, using "intensive support from a range of practitioners."

Grossman said that investigating the potential harm of child abuse prevention programs will also be important for future recommendations.

"We frequently think of removing (abused and neglected) children from the home as being a potential solution," he said - and often it is. "(But) removing children from the home and placement in foster care carries its own harms."