Patient Restraints: Bad Medicine?
More than 800 of every 1,000 hours psychiatric patients were hospitalized at a rural North Carolina hospital from July 1 through Dec. 31, 2013, were spent in some kind of physical restraint.
That works out to being restrained -- often bound by arms and legs -- for 33 days out of every 42. It's also more than six times the rate of the hospital with the next highest restraint rate and more than 800 times than the national average, which is less than an hour out of every 1,000. Those data reflect the most recent numbers from the Centers for Medicare and Medicaid Services.
Park Ridge Health in Hendersonville, North Carolina, is an outlier in an industry that 2 decades ago decided restraint caused more harm to patients than any potential benefits to their health or hospital staff. Even so, industry experts say restraining patients is much more common than it should be and often is a sign of overworked and understaffed nurses.
"You Don't Restrain Unless There Is No Other Way"
According to rules set forth by CMS, hospital staff may restrain patients only when medically necessary and to use the least restrictive form of restraint possible. The most restrictive form is 4-point restraints with leather straps, but hospital staff can use soft Velcros, roll-belts, or even sedative drugs to restrain patients. The data in this report do not not include restraint with sedative drugs, and is limited to psychiatric patients in general hospitals and freestanding psychiatric facilities.
"One of the things you're taught early on is you don't restrain unless there is no other way," said Jean Ross, president of National Nurses' United, the largest nursing union in the U.S. For instance, restraints may be used to prevent a patient from self-harm or harming others, she said, but they're to be used for as little time as possible and should be titrated to a less restrictive form of restraint.
When asked to speculate as to why Park Ridge relied so heavily on the use of restraints, Ross suggested the problem may be staffing. She said the hospital was short-staffed, thus there were no nurses or aides to work directly with the patients to calm them.
Patient restraint has changed a lot over the last 20 years, Boston University health law professor George Annas told MedPage Today/VICE News. Although restraints were once commonly used, the high profile Catherine Shine case as well as an investigative series by the Hartford Courant in the 1990s changed the way people viewed it.
In 1990, Shine's severe asthma prompted her to go to the emergency room at Massachusetts General Hospital. The 29-year-old arrived with her sister at 7 a.m., and she was given oxygen and medication through a nebulizer. The doctor told them he wanted to intubate Shine, but she repeatedly told him she did not want to be intubated, and said that she wanted to leave. The arguments continued, until her sister told her to "run" at around 7:40 a.m.
They were apprehended by doctors and a security guard and Shine was placed in 4-point restraints. By 8:25 a.m., Shine was forcibly intubated despite tests that showed her blood oxygen levels had improved with the oxygen and medication alone. She never gave consent, and she wasn't told about the risks and benefits of being intubated.
Shine was sent home the following day, but she died 2 years later because she had an asthma attack at home and was too afraid to go back to the hospital.
"I think it's just become recognized as just abusive, used for convenience of staff, not for welfare of the patient," Annas said. "There are hardly any medical indications for restraining someone. What could there be? It's much more an act of a jailer than it is of a physician or a nurse or a nursing assistant."
A Look At the Inspection Reports
Inspection reports obtained by MedPage Today/VICE News indicate that Park Ridge's restraint policies have repeatedly been called into question over the last 15 years. The reports, called CMS 2567s, are compiled by state officials and submitted to CMS.
In 2010, for instance, a man in his late 30s who spoke very little English was "roaming" around the emergency department, "initially looking for his brother" and acting "increasingly hostile," according to one of the inspection reports. After talking with police, he said he had pain in his "left chest," and calmed down.
The man refused behavioral medication, and the situation escalated. Four sheriffs, three nurses, and the physician restrained the patient face-down on the floor. They then administered an intramuscular shot and handcuffed him. Once the man was allowed to stand, they applied 4-point soft restraints, but there wasn't a time limit on those restraints, which is required and should be on no longer than 4 hours, according to CMS policy.
The patient had "cognitive impairment that interferes with medical care" and could remove medical devices, tubes, and dressings.
"Unable to follow safety instructions," the inspection report reads, quoting staff notes. "Less restrictive methods have been determined ineffective."
But investigators concluded that hospital staff did not attempt to use these "less restrictive methods" before allowing the sheriff's department to use handcuffs. The hospital's most restrictive option, called the "Dr. Strong," a code word for getting assistance from hospital staff trained in acceptable patient restraint, was not called before the patient was handcuffed. And hospital staff did not time-limit the restraint order or monitor the man in restraints, which they are required to do under federal policy.
Two years prior to the handcuff report, Park Ridge was inspected because a patient died in restraints. Although all patient deaths associated with restraints or seclusion are required to be reported to CMS, a 2008 inspection report reveals that Park Ridge didn't report a 75-year-old man with dementia who died while be was being restrained via a lap belt. (He also had other ailments, including coronary artery disease.) The day before, a doctor had written a 24-hour order for restraints. Inspectors also noted that Park Ridge did not maintain a log of deaths reported to CMS.
Another patient at the hospital in 2010 was restrained without a physician's written order to do so for 22 hours and 37 minutes, according to another report. At 22 hours, 38 minutes, the physician wrote an order. Not only is such an order required prior to restraint under federal law detailing patient rights, but an adult patient is not supposed to be restrained for more than 4 hours without an additional order, and the limit is not to exceed a total of 24 hours.
The same report chronicles other incidents in which patients at risk of falling went up to 13 hours without being monitored while in restraints. Hospital policy requires monitoring every 2 hours for these patients, according to the report. For one patient who wasn't monitored as regularly as required, the IV tubing and bags had "change by" dates up to 5 days prior to the date inspectors toured the facility.
Despite Transparency, High Rate Unknown Locally
Over the course of the MedPage Today/VICE News investigation, the North Carolina Nurses Association and Disability Rights North Carolina both said they were unaware of Park Ridge Health's high CMS restraint records, which have been public since at least 2014, when reporters in Connecticut spotted Park Ridge's high rate in 2013 data.
Down the street from Park Ridge Health, a retirement community called Fletcher Park Inn boasts of its location the hospital as part of its marketing: "Having a 103-bed faith-based hospital, operated by Adventist Health Systems, across the street at Park Ridge Health is reassuring to even the healthiest of our retirees."
Once she looked at the 15 years of inspection reports obtained by MedPage Today/VICE News, Kristine Sullivan, a lawyer for Disability Rights North Carolina, said the hospital was cited again and again for overusing roll belt restraints to prevent falls from 2002 to 2013, culminating in citations for abuse, neglect and even the death of a patient.
"As we have learned from our work in adult care homes and nursing homes, restraint use actually increases the likelihood of falls and can cause additional harm to patients who fall while in restraint," Sullivan said. "CMS has put a lot of information out about avoiding the use of restraints in response to falls."
Sullivan said it's important to remember that Park Ridge isn't the only hospital with patients that exhibit "undesirable behaviors." Across the state, a lack of sufficient community mental health services drives these patients to the emergency room, where they're sometimes met with hospital staff that doesn't know better than to restrain them for convenience, she said.
"That said, the sheer number of restraints that are used at Park Ridge suggest that this problem goes much deeper than a typical lack of knowledge or expertise," Sullivan said. "It suggests a systemic over-reliance on practices that are outdated and harmful; possible understaffing; and the need for a change in the hospital's treatment philosophy that must come from the top levels of management."
What Officials Had to Say
When asked about its unusually high restraint rate in the federal data, Park Ridge's chief nursing officer, Craig Lindsey, said the rate we found was an error, but that the hospital couldn't estimate what the correct number would be.
"In the Behavioral Health [unit], it is not uncommon to place patients in restraints at the start of their treatment, and then remove the restraints as we stabilize the patient," Lindsey said through a hospital spokeswoman. "The error resulted from a disconnect between the way that we were documenting the removal of restraints and where our [electronic medical record] 'looks' for the documentation of removal."
Lindsey also said the hospital had never been contacted by CMS regarding its restraint rate.
"We did not go back to recalculate the accurate number as we would have needed to dig through too many charts, and it was not worth the effort once we identified what the issue was," he said in response to a follow-up question.
When asked about the inspection reports, Lindsey said he couldn't comment because they happened before his tenure as chief nursing officer.
MedPage Today/VICE News also asked CMS about Park Ridge's restraint data, and why the high rate never prompted an investigation of some kind at Park Ridge.
Although MedPage Today/VICE News supplied the restraint data and inspection reports obtained via our public records request to a CMS official, the official did not answer questions about whether the rate was a cause for concern, whether it had ever reached out to Park Ridge regarding the rate or whether the data could be inaccurate.
Instead, the official said the restraint measurement is part of its Inpatient Psychiatric Facility Quality Reporting program, which is intended to encourage facilities to improve the quality of their inpatient care and disseminate information to the public so they can be better informed when they choose their health care providers.
The official said the inspection reports "speak to the situation at Park Ridge," and that, in general, hospitals with violations must submit plans of correction for CMS approval, and may be subject to a follow-up inspection to be sure they're in compliance.
Park Ridge Health spokesperson Victoria Dunkle stressed that the most recent CMS restraint data are from 2013.
"The data in question are outdated and based on an EMR reporting error calculated in 2013," she said in an email. "It is simply inaccurate. Park Ridge Health is confident the steps we have taken to address the issue will remove the disparity between our apparent use of restraints and the rates at other facilities."
Restraint data in the current database covers the final two quarters of 2013, but it was published in April of 2015. New restraint data for the entire 2014 calendar year is due out in April 2016. Again, it will include only patients in psychiatric hospitals and psychiatric units in general hospitals.
More Trouble for Park Ridge Health
Restraint investigations haven't been the only source of trouble for the rural hospital.
Last fall, Park Ridge's parent company, Adventist Health, reached an agreement with the U.S. Department of Justice to pay a settlement of $115 million stemming from charges it violated the False Claims Act by miscoding payments and having improper compensation agreements with physicians.
Several of the whistle blowers were at Park Ridge. One doctor made more than a $1 million in 1 year while only working 3 days a week, according to the report.
Other citations against the hospital over the years included repeated violations of emergency treatment requirements, which state that a CMS participant hospital must treat its patients or stabilize them before discharging or transferring them to another hospital.
"The most recent one involved discharging a 14-year-old pregnant girl who was in premature labor and led to [Park Ridge] coming very close to having their Medicare participation terminated," Sullivan said.
On June 6, 2012, CMS sent Park Ridge a letter stating that the deficiencies uncovered as a result of this incident were "so serious" that Park Ridge's emergency patients' faced an "immediate threat" to their health and safety. The letter went on to state that Park Ridge's Medicare provider agreement would terminate on June 29, 2012 if the hospital wasn't able to demonstrate compliance. The date was then backed up to July 6, 2012. CMS wrote on June 27, 2012 that it had received and accepted the hospital's plan to correct its deficiencies, stopping CMS from terminating the Medicare provider agreement.
Following the discharge of a pregnant teen in premature labor, CMS nearly terminated Park Ridge's Medicare provider agreement. The hospital wrote a plan of correction, which CMS accepted, allowing Park Ridge to keep its agreement, according to a June 27, 2012 letter.