Do No Harm

Do no harm,” an important phrase in the delivery of healthcare, is not working. In fact, depending on the epidemiological approach and which data set one applies, medical errors, hospital-acquired infections (HAIs), and pharmaceutical errors combined are the leading, second, or third killer of Americans and Canadians annually. According to Joe and Terry Graedon of the People’s Pharmacy in their new book, Top Screwups Doctors Make And How to Avoid Them (Crown Publishing 2011), medical mistakes are the leading cause of death when all the categories are counted, including: medical error, hospital-acquired infections, drug error, misdiagnosis, post operative infections, fatal drug reactions in Nursing Homes, unnecessary surgeries, and lethal blood clots in veins.


These errors, combined, amount to over 788,000 deaths per year—higher than heart disease (616,000) and cancer (562,000). According to the Centers for Disease Control (CDC), 2.4 million Americans died in 2007, making hospital intervention accountable for approximately one-third of all deaths in one year.


To get a truer picture of the scope of the problem is difficult as non-reporting rates are tremendously high—ranging from 60 to 95 percent. Added, as well, are the hundreds of thousands of patients who are harmed (morbidity), but not killed (mortality). Though the numbers must be understood in terms of the billion or so procedures that are done annually, still the existing numbers define an epidemic of harm. Some recent examples that made the headlines include:



“Hospital uses wrong kidney for transplant”
(Los Angeles Times, February 18, 2011)


“Surgical errors continue despite protocols”
(Archives of Surgery study as quoted by NY Times, October 19, 2010)


“Infection control lapses plague outpatient surgical centers”
(Seattle Times, June 9, 2010)


“Boy accidentally given double dose of medication at Tacoma Hospital” (Seattle Times, November12, 2010)


“X-Ray beam strays invisibly, harming instead of healing”
(NY Times,  December 29, 2010)


West Virginia Hospital Overradiated Brain Scan Patients”
(CT Procedures, NY Times, March 6, 2011)


As to the extent of the epidemic or errors, in the U.S., according to Joe and Terry Graedon:  

  • fatal drug reaction deaths in hospitals: 106,000 
  • fatal drug reaction deaths outpatient: 198,815 
  • fatal drug reaction deaths nursing homes:  41,652 
  • deaths related to misdiagnosis: 132,500
  •  healthcare acquired infection deaths in hospitals: 100,000 
  • deaths from infectious diarrhea in nursing
  • homes: 16,500 
  • deaths from excessive radiation from CT
  • scans: 29,500 
  • deaths from unnecessary surgery: 12,000 
  • deaths from surgical and post operative
  • complications: 32,591
  • lethal blood clots in veins (deep vein thrombosis and pulmonary embolism) leading to death: 119,000 


And in Canada: 


  • 220,000-250,000 hospital-acquired infections every year in Canada leading to 8,000-12,000 deaths (CUPE Report, 2011)

  • 141,000-232,000 adverse events per year in Canadian acute care hospitals causing 24,000 preventable deaths (Canadian Medical Association, 2004) 

  • of 3,000 Canadian patients surveyed 10 percent reported a medical error, 24 percent reported a poorly coordinated case (International Journal of Clinical Practice, July, 2011) 

  • 1.5 million medication errors annually 

  • In Quebec  from April to September 2011, 179,000 incidents including patient falls, medication error, and botched tests (Canadian Institute for Health Information) 

  • Reporting rates of medical error and adverse events range between 5 and 20 percent  

  • 1 out of 13 patients report an adverse event


The Systemic Causes vs. “Low Hanging Fruit”


Healthcare delivery has evolved improperly by building  constructs that lead to medical error. Systemic causes that are directly linked by the science to patient harm have been systematically ignored and in some cases even legally blocked—as in the attempts by many states to pass regulation to improve staffing ratios. “For profit care” (U.S.), factory medicine (Canada), accountability issues (cannot get healthcare workers to wash their hands between patients) hierarchies in human relationship systems (bullying), over-reliance on technologies, stress, working conditions, staffing ratios in both clinical and non clinical departments, legal issues that conflict with safety issues, cost benefit issues that put safety on the negative side (spending side) rejecting the savings side of the equation—all considered systemic causes, all lead to adverse patient events but rarely, if ever, are confronted or changed as they are deemed too expensive and too extensive.


One such systemic factor is the profit motive. The Journal of General Internal Medicine published a study in March 2000 entitled “Hospital Ownership and Preventable Events” showing that patients in for-profit hospitals are 2 to 4 times more likely than patients at not-for-profit hospitals to suffer adverse events such as post surgical complications, delays in diagnosis, and treatment of an ailment. The editorial in the Journal also cited that for-profit hospitals lowered costs by cutting nursing services. Additionally, for-profit systems for blood collection cost 5 to 15 times more to collect, with 1,000 percent more blood wasted and higher transfusion-related infections, such as hepatitis.


Medicine in Canada has built in causes for errors, as well. More patients per hour, more operations per day, less rest between cases, less time with each individual patient. Lower staffing also can occur in factory medicine paradigms leading to more error. Provinces in Canada are balancing budgets by cutting support and ancillary hospital department staff, despite studies that show infection rates can and do increase with decreased staffing. According to a CUPE Report, staffing of support personnel is being cut 10 percent in Canadian health-care facilities. Numbers of beds are also being eliminated, causing overcrowding.


In the U.S, the number of staff to patients is not at acceptable levels. Only two states have regulations calling for a ratio guideline—California and Washington. Lack of staff increases the potential for medical error and has even been cited in the scientific literature as a direct link to error and infection. For each additional patient over-assigned to an RN, the risk of death increases by 7 percent for all patients. Patients in a hospital with a 1:8 nurse to patient ratio have a 31 percent greater risk of dying than patients in hospitals with a 1:4 nurse to patient ratio. 


Longer shifts translate into higher numbers of medical error. Physicians in training who are scheduled to work long hours make 36 percent more serious medical errors with 5 times as many serious diagnostic errors. Fatigue-related error data is plentiful in the scientific literature. Fatigue-related preventable adverse events associated with the death of a patient increased by 300 percent for interns working more than 5 extended durations shifts per month.


Injury to healthcare workers (10 percent apply for workers compensation every year) contributes to medical error and compromises patient safety. Healthcare worker injury then becomes part of a loop that contributes directly to lack of staff per shift. Often an injured health-care worker is not replaced or is replaced with a per-diem, who is not as fully aware of methods and that, too, can contribute to medical error.


Poor working conditions contribute directly to medical errors as they have a negative effect on staff. With 62 percent of nurses leaving the profession because of the physical demands of the job, working conditions are contributing to both negative patient outcomes and national and state nursing shortages. In 115 studies included in a 2003 review, evidence was provided that working conditions affect patient outcomes related to patient safety, the rate of medication errors, and the rate of recognition of such errors after they occur. Working conditions can be defined as ergonomics, patient developmental flows, staffing, workload, scheduling, autonomy, etc.


Bullying, too, has a direct and indirect effect on medical error and negative patient outcomes. It especially impairs nurses in their cognitive effectiveness. “If a new nurse in your hospital saw a senior physician placing a catheter, but not complying with the (hospital’s) checklist, would the nurse speak up and would the physician comply? The answer is almost always, “There is no way the nurse would speak up.” What other industry would accept a routine safety violation that is associated with the deaths of tens of thousands of patients and not be held accountable?” (Learning Accountability for Patient Outcomes,” Pronovost, JAMA 2010).


A study of 1,700 nurses, physicians, clinical care staff, and administrators found fewer than 10 percent address behavior of colleagues that routinely includes trouble following directions, poor clinical judgment, and taking dangerous shortcuts. Specifically, 84 percent of MDs and 62 percent of RNs and other clinical care-providers had seen coworkers taking shortcuts that could be dangerous to patients. Fewer than 10 percent said they directly confronted their colleagues about their concerns and 1 in 5 MDs said they have seen harm come as a result. In one study, verbal abuse from physicians was noted by over 90 percent of participants and 76 percent witnessed negative nurse to nurse behavior. Nurses reported that 71 percent of those behaviors resulted in medical error, of which 29 percent resulted in death.


It must be noted that lack of real numbers hampers the research. The rates of under- and non-reporting are extremely high, running anywhere from 60 to 90 percent, depending on the study cited. There are 27 states in the U.S. with reporting regulations and none in Canada. But even in the states that have reporting regulations, there is a larger problem with compliance, funding, etc.


The legal system may be contributing to the overall problem of medical error. By not admitting error due to fear of liability and litigation, doing professional root cause analysis is compromised and therefore compromises care. There are also state and federal regulations that create legal mandates that can be in conflict with patient outcomes. A survey published in the Annals of Internal Medicine reported that nearly half of doctors say they have failed to report an impaired or incompetent colleague or serious medical error, though they were morally/legally bound to do so either by internal policy or state regulation. Accountability issues are constantly arising and being tested. Studies have shown that even getting health-care workers to wash hands between patients or after leaving bathrooms is not enforced and there are low compliance rates.


Interestingly, there has been a consensus that technology is the panacea. It is and is not. Computerized Physician Order Entry (CPOE) is being implemented at a huge financial cost. Smart technologies in health care are being designed to intervene in administration errors, including smart infusion pumps and bar code verification systems. But, according to a recent study, 98,000 people end up in emergency rooms every year (mostly elderly) due to medication error. Though new technology has shown good reductions of error, especially in the administration of pharmaceuticals, we must be careful not to adopt them as curealls, but more as support tools.


Attaching a cost-benefit analysis to medical error is a challenge, especially when health-care facilities do not understand the true science of the cost-benefit of medical error and many reject the premise of “indirect cost.” This can lead to miscalculations and decision errors. For example, the system sees expenditures on prevention as a cost. But the Society of Actuaries has stated that medical errors are costing 20 billion a year. Bed sores alone account for a cost of 3.9 billion annually. The cost per patient of medical error can be as high as $20,000 per bed (using the American Hospital Association’s data of one million hospital beds in the U.S.).


What Are The Problems?


One of the premises of this article is to look at the problem of medical errors not necessarily from a clinical perspective, but from a social science and political framework. This allows for a reasonable discussion of systemic causes, the concept in which health-care delivery is designed and the manner in which those design characteristics play a pivotal role in causing error and infection. Unless this type of analysis takes place, it is believed that the health-care community will miss the underlying reasons why error and infections are prevalent and only allow for the “low hanging fruit” (checklists, computerization, IT) to be selected for remediation.


The social sciences deal with the inter-relationships within a societal framework. A social science approach combines humanism, relativism, demography, communication, behavior (interactions), and other approaches in a broad-based theoretical analysis instead of just constructing empirically falsifiable theories. We believe that each health-care delivery system and its partner hospitals are a society within itself where departments interact much in the same way as different groups within a society interact. When dysfunction occurs there is a current that reverberates through the entire system of delivery of care.


The American Hospital Association, a group that is employer owned, made a laudable case and embarked on the “100,000 Campaign.” This shed light on a situation that was not on the national radar. Much was done and, in some cases, decreases in categories were achieved. However, the weakness of this effort was the voluntary participation of each individual health-care system designing its own definition of what is considered a medical error or infection. At present, 27 states in the U.S. have reporting regulations, but the compliance rates are abysmally low, under-funded, and few consequences were built into the regulations. There are no existing regulations in Canada, either federally or provincially, that require hospitals to report medical error or infection other than internal policies.


Changing systemically will be expensive, but the numbers actually speak for themselves—if they are disseminated impartially. If, for example, increasing staff would prevent X numbers of medical errors and/or infections, the dollar costs of expanding labor would be offset by decreasing costs of errors and infections. A bed sore can cost $14,000 per case. If we take that number, preventing three bed sores can pay for an extra full-time employee, which would theoretically prevent the bed sores by turning the patient more  often.


The Hippocratic Oath, “do no harm,” and the vision statements are posted at all hospital entrances. Yet, recently, a 40-year-old man was admitted for a routine shoulder procedure. After he was given too much anesthesia, he became oxygen-deprived and brain death occurred. In order to avoid the reporting regulation (reporting error within 24 hours of admission), the hospital put the patient on life support to keep him alive for one more day so the 24 hour reporting rule could be circumvented.


Compromises in patient safety have an ethical component. A 3 percent error rate might seem low, but then add a 3 percent hospital-acquired infection rate and a 3 percent drug error rate, and you have a 9 percent error rate—almost 1 out of 10 hospital patients. Patient safety committees, medical ethics committees, and risk management committees exist to deal with problems, but they are often functioning within existing frameworks of systemic dysfunction. They are then asked to break free of the constraints and pressures that are resistant to change during one-hour weekly meetings.


What are the performance achievements of healthcare in the U.S. and Canada? The U.S. ranks 20th in the world and Canada 9th in health rating inclusiveness, quality of service, and perceived health care between highest and lowest incomes. The Bertelsmann Stiftung Foundation of Germany in a 2011 report entitled “Social Justice in the OECD,” noted that in the U.S. has the most expensive care per person ($7,960) while Canada has the fifth most expensive ($4,363). Counter-intuitiveness aside, being an expensive health-care delivery system does not necessarily deliver quality health care and in the rush to treat as many people as possible, something gets overlooked in the Western medicine model.


Building The Case: A Temporary Conclusion


The social science perspective encourages the formation of a social movement on issues of such important societal impact as public health and national patient safety. A social movement would require health-care workers of all types, working with public health officials, legislatures, trade unionists, government agencies, and funding agencies, to write a plan of action to challenge the status quo of medical error based on the systemic causes listed here: staff ratios, shift work, bullying, overbooking and overcrowding, reporting errors, accountability where little or none exists, solving the contradictions between legal need, patient safety, and on and on.


The healthcare delivery system requires systemic change. We cannot even at this point count with any reliability on the number of errors due to lack of reporting, under-reporting, and self-definition reporting. There is no federal reporting rule and, in the states that have rules on reporting (only 27 at present), the language is weak and most states report low compliance. In Canada, there is no federal or provincial reporting rule at present for medical error. Hospitals are reluctant to release error and infection information. Hospitals that do report are guided by self-reporting standards that are unreliable for consumers. And then, of course, there is the fact that we are asking the people who are responsible for the very conditions that have produced this alarming threat to human life to fix the problem.


Health care is sick and needs doctoring. The damages being done  cry out for systemic change.




  • Britain in a report to Parliament 2008, 11,000 deaths per year due to medical error 

  • Germany, with 1/4th of the U.S. population reported 17,000 deaths per year due to medical error 

  • 15,000 Medicare patients die each month in part due to hospital care including such events as bedsores, thinners, infections and mis-medicating (DHHS, Office of Inspector General Report, quoted in U.S.A Today, Nov 16th 2010, p A1) 

  • Of the one million Medicare patients discharged each year, 134,000 were harmed by medical care and 180,000 die each year( Inspector General’s report, 2010) 

  • North Carolina study in 2008, 25.1 injuries per 100 patient admissions are injured due to medical error 

  • 1/3 of all patient admissions result in some form of medical error (Human Affairs, 2011) 

  • 40 wrong site surgeries occur every week in the U.S. (Human Affairs, 2011) 

  • Hospital deaths caused by MRSA in U.S. 1999-2001, sixth leading cause of death in U.S. costing 50 billion (Willems, G Emerging Infectious

  • Disease, volume 13 no 12, December 2007, 

  • 1.14 million medical errors out of 37 million hospitalizations in medicare population from 2000-2002 or 3.1 percent (Health Grades 2004) 

  • According to FDA data 1.3 million patients are injured per year from medication errors



William Charney, an occupational health scientist, is editor of Epidemic of Medical Errors and Hospital-Acquired Infections and The Handbook of Modern Hospital Safety.