Whenever a high-profile labor dispute comes to a head, those on the outside looking in are forced to sift through predictable posturing for actual facts.

But rarely are the stakes as significant and personal for the community at large as they are in the ongoing dispute between 14 Twin Cities hospitals and their 12,000 nurses. Minnesotans who depend on those hospitals for health care have a potentially life-or-death interest in the outcome of negotiations that restarted this week, at the request of a federal mediator, in an attempt to avoid a strike.

Patients and future patients will not have a seat at the table. They can only hope the two sides will do everything possible to reach an agreement on a new contract and avert a strike that would immediately test the quality of hospital care in the Twin Cities. As they negotiate over staffing ratios, pay and benefits, leaders on both sides have a responsibility that extends beyond boards of directors and rank-and-file membership.

Above all else, both sides claim patient safety is a priority. If that rhetoric can be believed, a chilling study published last March by the National Bureau of Economic Research provides an excellent reason for negotiators to find room for compromise.

Authors Jonathan Gruber and Samuel A. Kleiner analyzed data from 50 hospital strikes in the state of New York from 1984 to 2004 to assess the effects of strikes on two critical measures of health outcomes: mortality and readmission rates. They found that nurses strikes increase in-hospital mortality by 19.4 percent and 30-day readmission by 6.5 percent for patients admitted during a strike. Boiled down, that's one additional death for every 286 patients admitted during a strike, and one more readmission for every 111 patients.

"We show that this deterioration in outcomes occurs only for those patients admitted during the strike, and not for those admitted before or after to the same hospital," Gruber and Kleiner wrote. "And we find that these changes are not associated with any meaningful change in the composition of patients admitted during the strike or the treatment intensity for patients admitted during these strikes."

The authors also found evidence of more-severe ramifications for patients who required more-intensive nursing care, and that outcomes were no better in hospitals that were staffed with replacement workers.

Because hospitals are critical to the nation's health care system, and due to concerns about the impact of strikes, they were excluded from U.S. collective bargaining laws until 1974. Once free to unionize, they quickly became a key source of union jobs, and today more than 15 percent of hospital employees belong to unions, representing 6 percent of all union employees in the country. Contrary to overall trends, union activity is picking up in hospitals, Gruber and Kleiner point out, making their research even more relevant as hospitals face increasing pressure to cut costs.

Gruber and Kleiner say their study was an effort to determine if concerns about the short-term impact of strikes on patient care were justified -- not to analyze strikes' potential positive or negative long-term effects on the quality of patient care. In other words, some strikes could lead to agreements that produce long-term improvements in care.

The Twin Cities hospitals are quick to point out that a study of New York hospitals is not necessarily applicable to Minnesota, and they emphasize that they would continue to provide a high level of care in the event of a strike. But those assurances make the thought of a Twin Cities strike no more palatable.

The most convincing way for the hospitals and nurses to show their commitment to patient safety is to agree on a contract and avoid providing researchers with a Minnesota case study.