But more significantly, the number of doctors and nurse practitioners who reported receiving payments shrunk by 51%, from 204 to 99.

“I would assume it was the stigma,” said Dr. Arthur Gale, contributing editor at Missouri Medicine. “You can’t pick up a newspaper and not read about Purdue. Even the greatest promoter of OxyContin and narcotics, Dr. Russell Portenoy, is now saying he was exposed to false information.”

Data from the Centers for Medicare and Medicaid Services (CMS) show that a small group of doctors in Connecticut received the bulk of payments during the two years. The top five recipients are Dr. Thomas Pellechi ($747,429), an internist in Greenwich; Dr. Michael Brennan ($41,611), a physiatrist who specializes in rehabilitation medicine in Fairfield; Dr. Howard Hochster ($11,541), an oncologist in New Haven; Dr. Dean Mariano ($10,680), an anesthesiologist in Meriden; and Dr. Lucien Parrillo ($205), an internist in Portland.

The number of doctors and nurse practitioners receiving payments from Purdue dropped 51%.

Pellechi, employed as a company doctor for Purdue of Stamford while also maintaining a private practice, earned the most of any physician—a total of $1.877 million from 2014 to 2018—more than three-fourths of the $2.24 million Purdue paid out during that period.

“The only conflict of interest is if I overprescribed, and I barely prescribed OxyContin,” he said. A review of the Medicare Part D prescription database showed Pellechi was not listed as a provider. His position with Purdue ended in mid-2019.

Even though individual payments from Purdue dropped sharply, critics argue that some doctors, such as Brennan, took too much. Brennan was paid a total of $150,555 between 2014 and 2018.

“Doctors like Michael Brennan received larger payments because they are in a category called ‘Key Opinion Leaders,’” said Dr. Andrew Kolodny, medical and co-director of the Opioid Policy Research Collaborative at Brandeis University. Kolodny has been a paid expert witness in opioid lawsuits. “Doctors were hearing from [and influenced by] pain specialists like him with a thriving private practice.”

When contacted, Brennan directed an office staff member to refer any questions to a study on the risks of withdrawing from opioids rapidly.

The financial relationships between pharma payments to doctors and their prescribing habits have been scrutinized for years.

A June 2019 study of prescribers under Medicare Part D in the journal Addiction found that “physicians who receive direct payments from providers for opioid prescribing tend to prescribe substantially larger quantities, particularly for hydrocodone and oxycodone.” The study said that doctors who took payments from opioid companies prescribed 8,784 daily doses of an opioid per year in excess of prescriptions made by their peers, who did not receive such payments.

“Doctors like Michael Brennan received larger payments because they are in a category called ‘Key Opinion Leaders.’ Doctors were hearing from [and influenced by] pain specialists like him with a thriving private practice.”

— Dr. Andrew Kolodny

Based on an analysis of Medicare Part D prescriptions in Connecticut, the top recipients of Purdue money had the following prescription rates in 2017, per the latest data: Opioids consisted of 71.91% of Brennan’s 7,202 prescriptions overall; 62.39% of Parrillo’s 2,963 prescriptions; and 9.77% of Hochster’s 133 prescriptions.

Other high prescribers of opioids who also accepted money from Purdue include Dr. Mohan Vodapally (72.79% of 6,431 prescriptions); Dr. Anand Rahul (34.3% of 1,029); Dr. Robert Boolbol (52.13% of 3,401); Dr. Vincent Carlesi (46.25% of 2,536); and Dr. Igor Turok (46.89% of 2,730).

Not all doctors who were paid by Purdue were listed as prescribers of opioids in the Medicare Part D database.  Dr. Joshua Hurwitz, a reproductive endocrinologist in Danbury, for example, was paid $2,227 in 2014 for consulting work in the area of women’s health and hormones, but was not listed as a prescriber of opioids.  “This is an example of appropriate physician consulting that has nothing to do with the opioid epidemic,” said Hurwitz.

“What’s the difference between consulting and giving talks on behalf of certain drugs?” pharma critic Gale asked. “Doctors shouldn’t accept money from drug companies because it’s tainted.”

Purdue paid 563 prescribers small amounts ranging from $10.18 to $200.17 between 2014 and 2018, indicating a free meal at an event or lunch with a sales rep.

But there’s no such thing as a free lunch, Kolodny pointed out. “Medical literature show that these small payments influence doctors’ prescribing habits, and that’s why Purdue did it.”

Purdue announced in 2018 that it would no longer send sales reps to doctors’ offices. But state attorneys general, including William Tong of Connecticut, say the damage is done. A lawsuit by the state of Connecticut alleged that a patient taking the lowest dose twice a day for a week earned Purdue $38. But if the patient instead took the highest dose, Purdue made $210—an increase of 450%. To obtain that revenue, Purdue formulated its sales strategy to increase prescribed doses, the lawsuit says.

Purdue’s flagship drug, OxyContin, was launched in 1996 with the highly addictive oxycodone as the single active molecule.  Oxycodone has had devastating consequences in Connecticut.  According to the Connecticut Office of the Chief Medical Examiner, between 2012 and 2019, there were 707 accidental overdose deaths involving oxycodone, though the Office does not indicate how many, if any, of these deaths involved OxyContin.

Deaths from all opioids climbed to 5,790 over the eight-year period, the state reports.

An October 2019 study by Yale University and the University of Connecticut published in the journal Drug and Alcohol Dependence found that overdose deaths from opioid use in Connecticut doubled in the last six years as a result of fentanyl and polysubstance abuse.

Findings such as these could explain why a drop in prescriptions alone doesn’t tell the full story. “There are a lot of pathways to addiction,” said Dr. Gregory Shangold, an ER physician in Willimantic.

CORRECTION: An earlier version contained incorrect information concerning Dr. Joshua Hurwitz.  He was not listed as a prescriber.  He was paid $2,227 for consulting work in the area of women’s health and hormones. C-HIT apologizes for the inaccuracy. 

This story was revised from an earlier version.

Matthew Kauffman, a freelance data specialist, contributed to this story.

An analysis of state data in a national report by the Dartmouth Atlas Project also shows that Connecticut’s Medicare program relies heavily on brand-name drugs, versus generics, especially in wealthy towns in Fairfield County – a factor that could be contributing to the state’s ranking in the top 10 nationally in prescription drug spending per patient.

Connecticut seniors spent an average of $2,795 on medications in 2010 – 45 percent higher than the lowest-spending state, Minnesota, and the highest rate in New England.

The new report provides an in-depth look at how prescription drugs are used by Medicare beneficiaries, age 65 and older, in the program’s Part D drug benefit, which had 37 million enrollees in 2012. It shows wide variations in the use of both effective and risky drugs among the 306 regional health care markets across the U.S.

Jordan V. Harrison Graphic

While the underlying health status of populations is a factor in prescription drug use, “it really does not explain the variations in drug use intensity that we observed,” said Dr. Nancy Morden, a lead author of the study.

The geographic swings in prescription quantity and quality “suggest that there’s something in the regional practice culture, and perhaps in the patient culture, that is driving these patterns,” she said.

Data included in the report, from 2010 Medicare claims, shows that Connecticut fares better than average in prescribing “effective” drug therapy to patients with certain serious conditions, such as heart attacks and diabetes. More than 81 percent of patients hospitalized for heart attacks were continuing to receive the recommended beta-blockers seven to 12 months after a heart attack  – higher than the national average of 78.5 percent.

In the area of discretionary medications, Connecticut’s use of antidepressants for seniors was slightly higher than the national average – 19.1 percent, compared with 18.8 percent. The rates among hospital regions varied – from a low of 15.5 percent in Milford, to a high of 22.6 percent in Meriden.

Similarly, patients receiving care in Stamford and Greenwich had higher-than-average rates of prescriptions for newer sleep sedatives, such as Ambien – 11.1 percent and 10 percent, respectively, compared to 7.6 percent nationally. That was double the prescription rate for patients in Putnam and Derby.  By state, the use of so-called “sedative-hypnotic” medications by Connecticut seniors was the highest in New England, the data shows.

While the newer sedatives initially were considered safe, recent reports have shown they can cause persistent drowsiness, as well as other side effects that may be more pronounced in the elderly, Morden said.

The prescription rate for dementia drugs also varied, from a high of 9.3 percent in Meriden, to a low of 5.3 percent in New Milford. The statewide rate was lower than the national rate of 7 percent.

Generally, the use of “high-risk” medications in Connecticut also was lower than the national average.

The researchers said the regional differences in the use of discretionary and high-risk drugs, some of which have uncertain benefits, raise concerns.

“[The] regional variation highlights the absence of a ‘best practice’ consensus” for the drugs, the report says.

Overall, Connecticut patients filled a lower-than-average number of prescriptions in 2010 – 46.5 per patient, compared with 49 percent nationally. Still, the state’s costs per patient were high. By hospital area, Meriden had the highest drug spending — $3,248 per patient — while Winsted had the lowest spending– $2,354 per patient.

Morden said the research team found no correlation between higher spending and the rate of “effective” drugs being dispensed – dispelling the notion that higher spending means better care. Instead, spending is driven by the number of prescriptions – which were below average in Connecticut — and medication costs, she said.

“If you see high spending, without high quantities, it’s fair to assume that your prescribers are selectively using more expensive products,” she said.

Connecticut’s proportion of brand-name prescriptions, versus generics, was the highest in New England, at nearly 30 percent. All other neighboring states were well below the national average of 26.3 percent. By hospital area, Greenwich, Stamford and Norwalk had the highest brand-name usage rates, with Greenwich nearing 40 percent.

Branded drugs are generally more expensive than their generic alternatives, although they are therapeutically equivalent. Nationally, only a few communities, including South Miami, Fla., and Encino, Ca., had brand-name use rates as high as Greenwich.

For some illness types and severities, only brand names are available. But the Dartmouth team found that illness explained only 27 percent of brand-name use; the reasons driving the choice remain largely unknown.

Ellen Andrews, executive director of the Connecticut Health Policy Project, said differences in physician practice styles could explain some of the spike in brand-name use.

“There’s a lot of geographic variation in how doctors prescribe,” she said.

She also noted that prescription costs are just one piece of the health-care picture, which includes hospitalization rates and other measures.

While the study focuses on drugs, it shows that Connecticut’s Medicare spending on other services also is high. The state ranked ninth highest nationally in 2010 in non-prescription medical expenditures.

Morden said that generally, high use of brand-name drugs “doesn’t make any sense, clinically” and is likely driven by patient or prescriber preferences. Because the complex Part D Medicare structure leaves patients responsible for a significant portion of drug costs, income could be a factor in decision-making.

While the report did not single out states or regions, it found that higher spending in some areas was fueled by “greater use of brand-name drugs that in some cases may not provide significant additional benefits to patients.”

Nationally, the report found that seniors in the Miami region had the most prescriptions – nearly 63 per patient in 2010. Manhattan also was high, at 54 prescriptions.

Seniors in Miami also had the highest average spending on prescriptions, at $4,738; and the highest rates of at least one antidepressant prescription and dementia medication.

Manhattan held the top spot for sleep sedatives, with 15.3 percent of patients receiving a prescription.

The researchers said they hoped the report would prompt policymakers to take steps to equalize care, so that it is not a function of a patient’s zip code.

“Regional variation of the magnitude presented in this report . . . presents an opportunity for policymakers to study successful regions that provide effective care efficiently, determine what factors lead to this success, and disseminate these systems more broadly,” they said.