TOO MUCH DOUBT raised about its impact and too much division even among those who were supposed to be its bedrock backers.

That double whammy proved deadly for the campaign to mandate minimum nurse staffing levels in Massachusetts as voters soundly defeated Question 1 on Tuesday. Based on incomplete returns, the ballot measure was being rejected by a 70-30 margin. It marked the end of a highly-charged, often bitter campaign that saw millions of dollars spent by both sides on ads that featured nurses and relied heavily on frightening people into believing a victory by the other campaign could have life or death consequences for patients.

The measure was brought to the ballot by the Massachusetts Nurses Association, the state’s largest nurses’ union, which argued that inadequate nurse staffing in hospitals was endangering patient safety and care. The campaign against the question was bankrolled by the state’s hospitals, which argued that the it would drive up costs while imposing unwieldy, rigid staffing restrictions without improving care. Opponents argued that the measure would have further harmful effects on access to care at community hospitals serving lots of low-income patients by mandating new costs at already financially fragile institutions.

Proponents saw their lead in early polls slip away as public opinion turned sharply against the measure, which hospitals pounded at with advertising warning of its impact. The heavy advertising blitz came on top of an analysis issued last month by a state oversight panel that raised questions about the ballot measure’s cost.

“We are all disappointed by tonight’s results and the impact this will have on the patients we care for every day,” said Donna Kelly-Williams, president of the Mass. Nurses Association, who addressed reporters at the union’s Canton headquarters. “We know that right now – as I speak to you here – there are nurses caring for too many patients, and those patients are unnecessarily being put in harm’s way. And the problem continues to grow every year. The status quo is not a solution here.”

The Massachusetts Health & Hospital Association praised the result, but offered conciliatory words on the overall issue of hospital staffing.

“What we won tonight was the ability to continue providing the best possible care for patients throughout Massachusetts,” said Steve Walsh, president of the hospital association, in a statement. “This is the beginning of a conversation, not the end. Question 1 forced some difficult and necessary discussions about the future of health care and the future of our workforce going forward. These conversations with our care teams and in our communities have been critically important and will continue in bargaining sessions, legislative debates, board rooms and newspapers.”

Gov. Charlie Baker came out against the ballot question, while some of the state’s leading Democrats, including both US senators and Boston Mayor Marty Walsh, endorsed the yes campaign.

In the Legislature, neither House Speaker Robert DeLeo nor Senate President Karen Spilka staked out a position, though DeLeo’s top lieutenant, House Majority Leader Ron Mariano, penned an op-ed last week opposing the measure.

The ballot question would have limited the number of patients a nurse could care for, with standard hospital floors limited to four patients per nurse. For more acutely ill patients, nurses would have been limited to fewer patients.

In early October, the state Health Policy Commission released an analysis of Question 1 estimating that implementation of the ballot question could increase health care costs in the state by as much as $949 million per year and require the hiring of as many as 3,100 new nurses.

The report also looked at the experience with a nurse staffing law in California, the only state with mandated minimum patient-to-nurse staffing ratios, concluding that there was there was no clear evidence of systematic improvement in patient outcomes from the state’s implementation of the law in 2004.

The report was big blow to proponents of the ballot question, with its mixed assessment of the California law’s effect and its estimate of cost impacts that closely tracked a projection issued earlier by the state’s hospitals.

The ballot question campaign came after two decades of unsuccessful efforts by the nurses’ union to get a staffing ratio bill passed by the Legislature. The union was successful in 2014 in getting legislation passed to establish minimum staffing requirements in intensive care units, but efforts to expand such mandates to all hospital settings hit a dead-end.

When the MNA pivoted and launched its effort to put the question directly before voters, there was good reason for the union to feel optimistic.

The idea of requiring minimum levels of nursing care seemed like a question that would have broad appeal. What’s more, the state’s nurses, who would be the best ones to make the case in a grass-roots ballot question campaign, seemed fully on board. In April, the MNA released a poll of Massachusetts nurses showing that 86 percent supported nurse staffing limits.

In September a Suffolk University/Boston Globe poll had Question 1 on top, with 52 percent support and just 33 percent opposing it, while a WBUR poll at the same time had the question deadlocked 44-44.

By early October, however, the picture had reversed, with a UMass Lowell/Boston Globe poll showing 51 percent of likely voters opposing the question and only 43 percent supporting it. A WBUR poll in late October also showed the measure slipping badly, with the no side ahead 58-31.

A separate WBUR poll of nurses in mid-October showed what may have been the jarring poll result of the campaign: According to the survey of 500 registered nurses who said they were likely to vote, opinion was almost equally divided on the ballot question, with 48 percent saying they planned to vote yes and 45 percent saying they planned to vote no. It was a stunning shift from the overwhelming support among nurses reported in the spring by the nurses’ union.

Meanwhile, a late October Suffolk University/Boston Globe poll found that nurses had tremendous sway with voters, with 44 percent of those surveyed saying input from a nurse they knew was the most influential factor in their decision.

Pollsters and campaign strategists say opponents of ballot questions have a built-in advantage, especially in the case of issues as complicated as the nurse staffing measure.

“When there’s confusion and doubt, the status quo is the safe place to be, and that tends to be the ‘no’ side,” said George Cronin, managing director at the strategy and lobbying firm Rasky Partners and an expert in ballot question campaigns.

That was the case for Beverly Williams, a retired Boston public school teacher. “I was torn, but I ended up voting no,” Williams said at the Codman Apartments polling place in Dorchester. She said the issue should be worked out in contract negotiations between nurses and hospitals.

Cronin said the further doubt introduced by nurses appearing to be so divided on the question was particularly devastating for proponents. “I think the cautionary tale is, if you’re going to bring something to the ballot, you better be sure you’ve got a unified constituency and you don’t give the other side the ability to peel away your constituency,” said Cronin.

Chris Anderson, co-founder of Anderson Robbins Research, which did polling for the Mass. Nurses Association, said their opponents fully exploited the advantages of working against a ballot question. “They have run a very effective and ruthless textbook ‘no’ campaign,” he said.

The campaign was one of the most expensive ballot question showdowns in state history. The hospitals bankrolling the “no” side outspent the nurses’ union by more than 2-to-1. The “no” campaign reported spending $24.5 million, while the yes campaign spent $11.5 million.

The two sides battled to win over voters on a complex issue that frequently became the central topic of election conversations among Massachusetts voters leading up to Tuesday’s balloting. Sorting out the specifics of what the ballot question would mean, if approved, wasn’t easy. Staffing ratios, acuity scores, and the distinction between correlation and causation aren’t the stuff of bumper stickers or 30-second television ads.

The campaigns resorted instead to a combination of pulling at heartstrings and striking fear into voters. Proponents’ ads said patients were imperiled by low nurse staffing, while one “no” ad warned of a child needing emergency room care being turned away by a hospital in order to avoid violating its mandated patient-to-nurse ratio.

One Boston nursing workforce researcher slammed the ads run by both campaigns. “The advertising on both sides of these issues has been terrible,” wrote Karen Donelan, an associate professor at Harvard Medical School and senior scientist at the Mongan Institute Health Policy Center, in a commentary piece posted by WBUR. “In my opinion, both sides have stretched the truth and even flat-out lied about some of the research underlying their claims. Both have resorted to ugly scare tactics. Both sides should be ashamed.”