Dr. Albert B. Sabin, creator of the oral polio vaccine, went to Brazil in February at the "flowery" invitation of the country's health minister to help devise programs to wipe out the disease that annually cripples thousands of Brazilian children.

Now he is back in the United States, his task undone, disillusioned after more than two months of 12- to 16-hour days in seven-day weeks. The man who invited him, Health Minister Waldyr Arcoverde, and his chief deputy engaged in "untrustworthy professional behavior," Sabin charged.

Sabin had counted on Brazil to be as resolute as Cuba was when it eradicated polio 17 years ago. But though he told Brazilian President Joao Baptista Figueiredo that he would have liked to go on working "to eliminate preventable suffering and death," he said yesterday that he left Brazil "because I found it impossible to continue collaborating" with Arcoverde and his deputy.

From the start, they were unwilling to work together to do what had to and could be done eliminate polio, Sabin, 73, said yesterday in a telephone interview from Charleston, S.C., where he is distinguished professor of biomedicine at the Medical University of South Carolina.

The episode has been widely publicized in Brazil. Joined by organizations of physicians and statisticians, newspapers have praised Sabin and strongly criticized Arcoverde, in some cases calling on him to resign.

Over a period of 19 years, the scientist has become well-known in Brazil. His wife is Brazilian and he often has been consulted by the country's health officials about polio.

At the Brazilian Embassy here a news spokesman told a reporter that he would seek comment on Sabin's allegations.

Sabin drew a sharp line between the two top national health officials and state health officials. Many of the latter were excellent, he said. "We demonstrated for the first time that by proper organization it is possible quickly to stop epidemics of polio in countries like Brazil by the same rapid mass vaccination procedure with oral vaccine that was found to be effective in the United States 20 years ago."

Sabin's charge of untrustworthiness centers on the need for reliable data on the rates at which polio has occured and is occurring among children, so that realistic preventive and emergency-treatment programs can be set up.

In the five-year period starting in 1980, when polio first became a reportable disease in Brazil, the ministry of health told the World Health Organization (WHO) that the number of cases averaged 15,000 annually.

The figures were gathered by a government statistical agency whose employes, at the end of each year, visited hospitals in each of Brazil's more than 4,000 counties to find out how many polio victims had been admitted.

For the period 1973 through 1976, however, the ministry told the WHO that the number of cases had dropped a "remarkable" 86 percent to 90 percent a year, Sabin said.

Sabin discovered this discrepancy while in the capital city of Brasilia, and determined that the drastically lower figures derived from estimates compiled by the ministry from reports by state health departments. These reports, he said, were "totally misleading."

The switch in reporting to the WHO was made by "a couple of bureaucrats" -- epidemist who apparently acted without the knowledge of their superiors, Sabin said.

As a result of his discovery, Sabin pressed for a survey of the kind that has found in the last few years, high levels of previously unreported residual crippling polio in school children in several African and Asian countries.

Sabin's proposed survey would have involved about 1.4 million children, most of whom live in what he called "the most primitive conditions imaginable." m

Sabin also recommended modifications in a mass-immunization program to which the government already had committed itself.

The recommendations were laid before President Figueirdo at an early meeting with Sabin and Arcoverde. The president "became very, very enthusiastic about my working with the minister," Sabin said.

Presidentially blessed, Sabin went on to investigate conditions in four states -- Goias, Santa Catarina, Parana and the Federal District. He found that epidemics, prolonged by the absence of cold weather, had been going on for six to eight months.

Sabin and his associates mounted a mass vaccination campaign, similar to those first used in the United States, and stopped the epidemic in a matter of weeks, despite Brazil's much more difficult environment.

But in the end, despite the initial enthusiasm of the president and the health minister, Sabin saw his bright hopes dimmed.

A pilot survey of 25,000 school children in the Federal District was completed, but the proposed national survey foundered, Sabin said, after he told the health minister of the switch in the data base used to report polio to the WHO.

Then, in a March 10 memo to Arcoverde, Sabin told the minister that the persons he had chosen to head the national immunization program -- the epidemiologists who had provided the misleading figures to the WHO -- were not competent to head the program.

The combination created a situation, Sabin said, that made it impossible for him to remain in Brazil, despite his successful demonstration that the crippling disease could be wiped out there.

"I decided to leave," he said.