The Food and Drug Administration must bless any new drugs as “safe and effective” before they wind up in pharmacy aisles or prescribed to patients. But the ways in which the agency arrives at those approvals “vary widely in their thoroughness,” according to an analysis by researchers at Yale University’s School of Medicine.

“Not all FDA approvals are created equally,” said Nicholas Downing, lead author of the study, which examined nearly 200 new drug approvals between 2005 and 2012.

Researchers found broad differences in the data it took to get a thumbs up from FDA. For instance, the agency required that many new drugs prove themselves in large, high-quality clinical trials. But about a third won approval on the basis of a single clinical trial, and many other trials involved small groups of patients and shorter durations. Only about 40 percent of approvals included trials in which the new drug was compared with existing drugs on the market.

According to the Yale analysis, publicly available documents summarizing the FDA’s rationale for approving each drug showed the disparity in the quality of information and outcomes the agency required for each application.

“There was a lack of uniformity in the level of evidence the FDA used,” Joseph Ross, assistant professor of internal medicine at Yale and a senior author of the study, said in announcing the findings.

FDA officials said Tuesday that while the agency often tailors clinical trials and other requirements depending on the drug and the disease it is intended to treat, the standard for approval does not change.

“The agency applies the same statutory approval standards of safety and efficacy to all drugs, but uses regulatory flexibility in applying those standards,” spokeswoman Stephanie Yao said in a statement. “Some drugs may be tested in clinical trials that enroll hundreds of participants while others, particularly those seeking to treat rare diseases, may be tested in trials that enroll only a handful of participants.”

She added that the FDA maintains “a robust program for postmarketing surveillance to help ensure the benefits of marketed drugs continue to outweigh their risks.”

The Yale study, funded by the Pew Charitable Trusts and published Tuesday in the Journal of the American Medical Association, acknowledges that the FDA has valid reasons for its varied requirements.

“Such regulatory flexibility allows for a customized approach to approval,” it states, “including the ability to rapidly approve potentially effective therapies for life-threatening diseases, such as certain cancers, or those diseases for which there is no existing effective treatment, such as orphan diseases.”

Downing, the study’s lead author, said the analysis is not intended to suggest that the FDA should use a one-size-fits-all approval process or that it somehow has approved the wrong drugs. Rather, he said, it was meant to educate people who assume that all new drugs have cleared the same amount of safety testing before ending up on the market.

“It’s very understandable that regulators have a flexible standard for approval, but patients and doctors need to be aware the standard is flexible,” he said. “It’s important for doctors and patients to have a discussion about how much we know about the potential benefits or risks of taking a drug before they take it . . . [and] it underscores the need to continue to study these drugs after approval.”