Decisive help for teen mental health care.

The escalating rate of psychiatric hospitalization among adolescents has created intense controversy among insurers and mental health workers in the past decade. A federal review issued last month asserts that private psychiatric hospitals often inappropriately admit and hold people who have generous government insurance, many of them teenagers of military families.

However, psychiatrists can reach a surprising degree of agreement concerning which teenagers to hospitalize if they follow guidelines generated through a recently developed decision-making method, according to a report presented at the annual meeting of the American Psychiatric Association in Washington, D.C., last week.

"There have been egregious abuses of hospitalization for adolescents, but our data show that clinical decision making in psychiatry is not fuzzier or harder to agree on than in the rest of medicine," asserts study director Gordon D. Strauss of the University of California, Los Angeles.

Strauss and his colleagues relied on a method developed at the Rand Corp. in Santa Monica, Calif., and applied by other investigators to medical procedures in the 1980s. Strauss' team first devised a list of possible indications for hospitalization among adolescents suffering from substance abuse, conduct disorder or both. Psychiatrists assign these diagnoses to a large majority of teenagers admitted to psychiatric hospitals.

Indications related to hospitalization for conduct disorder include recklessness, suicidal or homicidal tendencies, positive or negative response to previous hospitalization, strong or weak family support, availability of outpatient treatment and presence or absence of other psychiatric problems. Many of the same indications apply to substance abuse, along with behaviors specifically linked to drug dependence.

Researchers divided each indication into high, medium and low levels, such as high recklessness and medium family support. Combining the indications in all possible ways for each diagnosis separately and together yielded 1,954 theoretical cases for evaluation by an admitting physician.

Nine national authorities on adolescent psychiatry, medicine and substance abuse then independently rated, on a nine-point scale, how strongly they felt each case needed hospitalization. General agreement emerged on 28 percent of the cases and substantial disagreement on 11 percent; the rest of the cases received equivocal or uncertain ratings.

Each panelist then reviewed the group's ratings and met with the others to discuss decisions and revise indications based on clinical knowledge.

In a second round of ratings, which expanded to 2,605 cases because of a couple of added indications, agreement on the need for hospitalization reached 55 percent and disagreement dropped to 5 percent.

Only past medical panels evaluating the need for tonsillectomy and adenoidectomy -- and dealing with fewer than 200 hypothetical cases -- have exceeded the level of agreement attained in the current study, Strauss says. Much lower agreement occurred among panels dealing with coronary artery bypass surgery and hysterectomy.

Disagreement regarding the need for psychiatric hospitalization of adolescents fell below that achieved for any medical procedure, Strauss remarks.

His group is using data from the study to develop software to aid in evaluating teenagers for psychiatric hospitalization. Its first customer, says Strauss, will be a "managed care" outfit that commissioned the study and oversees insurance payments for mental health care provided by some companies.

Further studies must establish whether psychiatrists can agree on the proper length of hospital stays for teenagers with substance abuse and conduct disorder, he adds.