Anyone who has experienced the chaotic environment of a hospital has an acute sense of the many things that can go wrong. In a typical visit, most patients encounter dozens of small oversights — from a misspelled name on a medical chart to a misscheduled diagnostic test.

For Rupinder Pannu, the error was more serious. She left the delivery room of the Trillium Health Centre in Mississauga, Ont., after giving birth to a baby boy in July 2008 with an object inside her that shouldn't have been there.

"The second week, I start smelling myself as a dead fish," she told CBC's the fifth estate. "My youngest son would come over and sit on me and say, 'Mommy, you smell funny.' "

The smell, it turns out, was coming from a gauze sponge that had been left inside her after she underwent an episiotomy, a procedure sometimes done during childbirth in which an incision is made in the perineum.

It was Pannu herself who found and removed the sponge after weeks of enduring pain so bad that she said she found it hard to walk.

"I would just sit in the bathtub with Epsom salts so my wound would heal faster, but that didn't help," Pannu said. "People would come visit me, and I would feel very stressed out [because of the smell]."

No compensation

According to Pannu, she only discovered the sponge when she decided to examine herself after the antibiotics a family doctor had prescribed, thinking the stitches used to close her incision had gotten infected, didn't get rid of the pain or the odour.

The hospital and the obstetrician who performed the delivery and episiotomy, Dr. Dalip Bhangu, met with Pannu after she went to the emergency department and reported what she had found. Pannu said hospital administrators apologized but told her she was not entitled to financial compensation.

Pannu complained about the incident to the College of Physicians and Surgeons of Ontario. In a 2009 letter responding to the complaint, Bhangu said he felt sorry Pannu had to "endure discomfort" but that when he left the delivery room, the nurse, who is responsible for verifying that all instruments, needles and sponges are accounted for, assured him the sponge count was correct.

He also said in the letter that he had been called to another delivery a mere 10 minutes after Pannu gave birth.

"I definitely hold the doctor responsible, because he is trained to do these things, and he cannot ignore me because he's got another patient," Pannu said of Bhangu's explanation. "If he doesn't want to work on me, he could just leave me rather than just leave stuff inside me and kind of walk on to another one."

According to a claims management company representing the hospital's insurer, neither the doctor nor any of the staff present during the delivery are to blame for the mistake.

"We do not believe that there is any responsibility for this unfortunate incident that rests with the hospital and/or staff," the adjuster, Cunningham Lindsey, wrote in a letter dated July 10, 2009, preemptively informing Pannu that any claim for compensation would be denied — even though Pannu hadn't made any official claim.

The hospital refused to comment to CBC News on the Pannu case and said only that "any error that impacts patients is investigated and lessons are developed from the incident and built into action plans for patient safety improvement." Dr. Bhangu declined to comment on the incident.

Surgical errors common

Mistakenly leaving foreign objects in a patient's body is one of the most common errors that occurs during surgeries, followed by operating on the wrong body part. It's one of the reasons why hospitals have adopted surgical safety checklists, intended to ensure that doctors and nurses follow a set of standardized steps before and after each procedure.

si-errors-instrumentsrtxy5j

Keeping track of the instruments used during a surgery is an important part of preventing errors. Hospitals have adopted something called a surgical safety checklist to help doctors and nurses do just that. (Pascal Lauener /Reuters)

Studies suggest surgery accounts for 40 to 50 per cent of all hospital-related adverse events, a catch-all term used to describe unintended injuries, complications or death related to the care received, not a patient's medical condition.

The checklists, devised by the World Health Organization in 2008, have been shown to reduce surgery-associated complications and deaths by more than a third and have been endorsed for use in birthing units by the Society of Obstetricians and Gynecologists of Canada.

The hospital where Pannu had her baby has been using surgical checklists since 2009 and in a statement said it uses two-person teams to perform supply counts and quality checks in its obstetrics units.

Studies spur change

Ever since the 1999 Institute of Medicine report To Err is Human raised awareness of the high incidence of preventable medical errors in the U.S., hospitals in many parts of the world have been trying to change what had long been a duck-and-cover approach to medical mistakes.

In Canada, the seminal study on hospital medical errors is a 2004 paper by Ross Baker, a professor at the University of Toronto's Institute of Health Policy, Management and Evaluation.

It found that 7.5 per cent of patients admitted to acute care hospitals in Canada in 2000 experienced one or more adverse events, which can include everything from reactions to wrongly administered medications to bed sores, falls, infections and surgical errors. Most of these events did not result in any serious harm, the study found, but almost 37 per cent were preventable.

More recent studies have shown rates of adverse events in hospital between 10 and 14 per cent.

"The groundbreaking nature of Dr. Baker's work in Canada and that of some of his colleagues internationally was to bring it out into the light of day," said Deb Jordan, executive director of acute and emergency services for Saskatchewan Ministry of Health.

No consistent reporting

Saskatchewan was one of the first provinces, along with Quebec, to introduce legislation in the early 2000s obliging hospitals to report critical incidents, adverse events that result in serious harm or death.

Simple solutions

Reviewing incidents and spotting patterns of dysfunction is a key part of any attempt to reduce medical errors.

"If I were king of our hospital, I would stop the whole place for a day and ask everybody to send me a list of their workarounds — all the ways in which they've found the system is a problem for them, and they've figured out a workaround — because those things are often the source for something falling through the cracks later," said Richard Boothman, executive director of clinical safety at the University of Michigan Health System in Ann Arbor.

When his risk management department examined the errors occurring in the health centre's oncology division, it found that 79 per cent of them were related to nursing — things such as falls and the misprogramming of pumps that deliver medications — so the unit boosted its nurse-patient ratios.

Other hospitals have analyzed medication errors and found they often occur when nurses are interrupted so they had nurses start wearing brightly coloured vests when giving out medications to signal they should not be disturbed.

"Often, these solutions, they're not high tech," said Wendy Levinson, chair of the department of medicine at the University of Toronto. "They don't need a lot of money to fix. They're human problems."

Most hospitals require critical incidents to be reported at least to the hospital board and the affected patient or family members, but such events represent only a fraction of the mistakes and close calls that occur in any hospital.

In general, Canadian hospitals have only a very vague idea of how many errors are made in the course of delivering care because there is no standardized system for measuring and reporting them.

Many hospitals rely on voluntary reporting of errors by staff, which has been shown to capture only a tiny proportion of errors.

"For most of our hospitals, the amount of staff that are dedicated to supporting patient safety is limited, and they often wear other hats and manage competing responsibilities," said Pat Campbell, president and CEO of the Ontario Hospital Association. "So, we could probably be making more rapid progress if we had more resources at each institution."

Some provinces, such as Ontario, require hospitals to report certain so-called patient safety indicators such as rates of hospital-acquired infections and patient deaths.

In 2002, Quebec became the first province in Canada to require mandatory reporting of what it calls "incidents and accidents," but it only began tracking them through a central online registry in 2011. 

Medication-related errors, which are the second-most common medical mistakes after surgical errors, are tracked federally through the National System for Incident Reporting, but reporting is voluntary (although Ontario requires hospitals to report to the registry).

"We don't have a 'Canadian health care system'; we have a series of provincial and territorial systems," said Hugh MacLeod, president of the Canadian Patient Safety Institute, which has established guidelines for how hospitals should disclose errors to patients.

"So, our desire is that people disclose [errors to patients], people report, and we're not too fussed about the mechanisms that they're using. What's important to us is that they're reporting."

Voluntary reports catch less than 15% of errors

But in fact, despite the widespread adoption of "no blame, no shame" policies, health practitioners are not reporting errors as often as they should be — usually because they fear repercussions or have misperceptions about which incidents should be reported.

Jurisdictions like Quebec, Saskatchewan and Manitoba that have been tracking critical incidents for years have found that while the number of incidents being reported has risen since reporting became mandatory, it still remains far below what research indicates it should be.

And without a systematic way of spotting errors, it's likely to remain that way.

A 2012 study by the inspector general of the U.S. Department of Health and Human Services found that voluntary reporting caught only 14 per cent of adverse events suffered by Medicare patients in U.S. hospitals.

But a method called the Global Trigger Tool, which relies on systematic reviews of patient charts by at least two health care professionals, was able to catch 90 per cent of the errors.

"A trigger could be a notation indicating, for example, a medication stop order, an abnormal lab result, or use of an antidote medication," the study authors wrote of the tool, developed by the Institute for Healthcare Improvement in the U.S. "Any notation of a trigger leads to further investigation into whether an adverse event occurred and how severe the event was."

A flaw in the system

Getting hospitals to examine the processes that lead to errors and near misses is a huge challenge, one that countries like Australia and New Zealand have been better at meeting than Canada, says Wendy Levinson, chair of the department of medicine at the University of Toronto.

'Systems need to be redesigned in order to help individuals in the systems do better, not just try harder.'— Wendy Levinson, chair of the department of medicine at the University of Toronto

"We have very much a philosophy inside medicine of 'We will try harder'; learning is trying harder to do it better," she said. "But what we haven't inculcated until recently is that instead of individuals trying harder, systems need to be redesigned in order to help individuals in the systems do better, not just try harder. It's a big distinction for us."

One hospital that has dramatically redesigned its approach to medical errors is the University of Michigan Health System (UMHS) in Ann Arbor, which includes three hospitals and dozens of clinics and care centres.

"I knew that we were not learning any lessons from what was happening to our patients," said Richard Boothman, a trial lawyer who instituted the overhaul in 2001.

"I had represented hospitals for 20 years in Michigan and Ohio and not a single hospital asked what we should have learned from the cases I handled."

Boothman, the UMHS's head of clinical safety, put in place a new system of reviewing patient charts and getting each clinical service to report regularly on a series of patient safety indicators specific to their department.

"We asked every single clinical service to give us the 10 things that if you heard them happening in your service, you would have to at least raise an eyebrow about the quality of the care," Boothman said.

Red flags include things like the number of emergency department patients who end up in intensive care after being admitted to a ward or the percentage of surgical patients who are back in the operating room with 72 hours.

Voluntary reporting has increased from 2,400 incidents in 2006/07 to 20,000 last year, Boothman said.

Full disclosure

UMHS also adopted a full disclosure policy when telling patients about errors and changed its malpractice strategy from an adversarial "deny and defend" approach to one in which it tries to resolve cases without going to court.

The health centre now preemptively offers patients financial compensation when it feels the standard of care has not been met, a method pioneered by the Veterans Administration Medical Center in Lexington, Ky.

The strategy has significantly reduced UMHS's malpractice costs, cutting the average cost per lawsuit by nearly a half, and decreasing the number of claims that wind up in court.

A similar compensation strategy is used by the Winnipeg Regional Health Authority — although it has not been as frequently applied as the Michigan model.

It's a brave approach, says Baker, the author of the 2004 Canadian study on adverse events.

"Not many organizations want to do that," he said. "It's an admission that we're not perfect."

Posting error data online

Part of changing the instinct to hide or contest errors is a willingness to disclose mistakes not just to patients but also to the public, and more and more hospitals are doing so on their own websites.

Montreal's Jewish General Hospital was the first hospital in Quebec to start posting information about adverse events online in 2011, following in the footsteps of what the Ontario Ministry of Health had started doing a few years earlier. 

"The initial reaction was a lot of concern from other institutions because it's not a requirement in Quebec for us to do this, so it wasn't something that individuals believed was the way to go," said Markirit Armutlu, co-ordinator of the hospital's quality program. "There was a lot of resistance because of fear of media reaction and public reaction."

Since then, other hospitals have come around to the idea and have started posting their own medical error data online.

Most don't reveal specifics but give only annual tallies divided into broad categories such as "medication," "falls" or "equipment-related" (though the Winnipeg Regional Health Authority reveals some details in the "learning summaries" it posts online).

Giving a public accounting of errors is a sign hospitals are being more accountable but doesn't necessarily mean they're any safer, warns Baker.

"The critical measure is how many changes are we making in the system as a result of what we're learning about these events," he said. "Saying how many reports we have is sort of like saying how many speeding tickets you got, not how many accidents you got."

To contact the Rate My Hospital team with tips or information related to the series, please email ratemyhospital@cbc.ca.

With files from Andrew Culbert