News

For the first time, a national study has tackled the contentious and provocative issue of death rates in hospitals, listing Canadian institutions by name and revealing where patients are likely dying avoidable deaths.

Though the Canadian Institute for Health Information researchers say the Hospital Standardized Mortality Ratio study should not be used to compare hospitals, it is inevitable.

By province, hospitals in Newfoundland and Labrador and Nova Scotia had some of the highest in-hospital mortality while many hospitals in New Brunswick, Alberta and Saskatchewan boasted some of the lowest.

In Canada's biggest city, the numbers are so dramatic that ailing patients in search of a specialist referral may wish to consider taking the subway downtown to Toronto's University Health Network, which not only sees the sickest patients but also had one of the lowest ratios. The same goes for St. Michael's Hospital, which also had low mortality figures.

The Hospital Standardized Mortality Ratio (HSMR) examines the numbers of expected deaths versus actual deaths in 85 acute-care hospitals at 92 sites across Canada. Of those, eight hospitals had some or all of their results suspended from publication, after hospital officials argued successfully issues of data quality. Other, smaller hospitals are also using the measure but were not included in the published report, due to small patient numbers.

The measure looked at 65 diagnoses that account for 80 per cent of in-hospital deaths including the top five: heart attack, heart failure, pneumonia, chronic obstructive pulmonary disease and septicemia, which is blood poisoning.

Data were adjusted for age, gender, urgent versus planned hospital admissions, whether patients were transferred from another hospital, the length of stay and other health problems, to name a few. In making those adjustments, researchers sought to make a level playing field for all hospitals.

Each hospital was graded against the Canadian average, which was assigned an HSMR of 100. A ratio equal to 100 suggests there is no difference between a local mortality rate and the average national death rate. A figure greater than 100 means a higher mortality rate, while below 100 spells a lower mortality rate.

Cape Breton Healthcare Complex in Nova Scotia fared the worst, with the highest overall figure of 137 from 2004-2007. Conversely, Regina General Hospital had the lowest figure - 71 - suggesting far fewer deaths than expected.

Nova Scotia Health Minister Chris d'Entremont conceded yesterday that the numbers are "particularly bad in Cape Breton," but said his ministry needs to look at exactly what the figures mean.

"We don't know if it's quality of care or just the kind of patients we have. Is it an issue with funding? I don't think it's to do with funding. It's to do with the aging and infirmity of the patients we do have."

Yesterday, Phil Hassen, chief executive officer of the Canadian Patient Safety Institute, said: "The big thing is to help people understand that there is work to be done. We have adverse events occurring and people are dying, but some of these can be avoided."

Each year, an estimated 23,000 Canadian patients die avoidable deaths in hospital, Mr. Hassen said, relying on figures in a different study. Even that number is conservative, he said, given that 8,000 to 12,000 of those casualties are due solely to hospital-acquired infections.

And yet efforts to reduce injury and death are not a secret to hospitals. Areas where the biggest reductions can be made to whittle down death rates and avoidable injuries include introducing rapid-response teams, ensuring best-practice management of heart attacks and cutting down on medication errors.

Though yesterday was the first public announcement of the HSMR data, it has been used by hospitals since 2004-2005. And since then, it has resulted in a 6-per-cent drop in standardized mortality over the past three years, which translates into 2,500 fewer in-hospital deaths, said Jennifer Zelmer, CIHI's vice-president of research and analysis.

However, researchers cautioned yesterday that the public should not use it as a shopping guide, saying patients with heart attacks or some other emergency should go to the closest hospital.

"The best will in the world and excellent care will not prevent all deaths," said Glenda Yeates, president and chief executive officer of CIHI. "Nevertheless, whether as a result of specific interventions or broader action, experience suggests that hospital death rates can change and improve."

While new to Canada, the measure has been adopted in many parts of the world since Professor Sir Brian Jarman of Imperial College in London, England, came up with the approach at least six years ago. It has been used by hospitals in England, Holland, Sweden and the United States, even forming the basis of campaigns to save patient lives.

At yesterday's news conference, Michael Baker, physician-in-chief of University Health Network, said the risk of medication errors has "dropped like a rocket" after the introduction of a special computer program that essentially double-checks the drug concentrations, volumes, types and doses of drugs.

UHN hospital president and chief executive officer Bob Bell said his network, which includes the three hospitals of Princess Margaret, Toronto General and Toronto Western, also introduced critical-care teams that quickly get lifesaving care to patients who are failing. The network's overall, three-year death rates are far below the average, with an impressive figure of 85.

Across the street from where Dr. Bell was being interviewed at Toronto General Hospital, it was a different story.

Mount Sinai Hospital, which has a terrific reputation, had a HSMR ratio of 111, which means it has higher deaths than expected. Maureen Shandling, senior vice-president medical, stressed that the number was within the range of other acute-care hospitals in Canada and that the "single number" doesn't reflect everything the hospital does.

"We provide excellent care in this hospital and we stand behind that care," Dr. Shandling said in a telephone interview yesterday. She added that the hospital has a quality and safety committee, excellent infection control practices and rapid-response teams.

Certainly, the report was not without its critics. Gordon Guyatt, professor of medicine at McMaster University in Hamilton, said the HSMR measure is far from foolproof, saying coding issues may play some part in low or high figures. If a patient, for example, had a heart attack and developed an infection and septicemia, does a health-care worker code it as a heart attack, sepsis or both?

"What additional factors did the patient have? High blood pressure when they came in?" Dr. Guyatt asked. "What is recorded and what is not recorded? The system is also vulnerable to factors that get into the medical record and the way they are coded."

Dr. Guyatt said that if a particular hospital has a high standardized mortality rate, it is not necessarily a reason to be terribly upset, nor if an institution has a low one is it a reason to celebrate.

Indeed, Ward Flemons, vice-president of quality, safety and health information for Calgary Health Region, had very impressive overall numbers for the region's three hospitals of Rockyview General Hospital, Peter Lougheed Centre and Foothills Medical Centre, with HSMR figures of 94, 88, and 84 respectively.

Still, Dr. Flemons was modest about the region's performance.

"There's a tendency for people to want to draw the conclusion that a low number must mean much higher quality of care and a high number must be the opposite," he said. "That may or may not be an accurate conclusion to draw and I don't think anybody knows one way or the other."

Despite the low numbers, he said it is not a cause for complacency, saying that "the value [of HSMR] is that it gets people talking and asking questions and I don't think that's a bad thing."

With reports from Avis Favaro, CTV News and The Canadian Press

THE LOWER, THE BETTER

A report graded each hospital against the Canadian average mortality rate, which was assigned a HSMR* of 100. A ratio equal to 100 suggests that there is no difference between the local and national average death rate. A higher rate suggests more deaths than average.

By score:

Of the 92 hospital sites in the study...

43 had fewer deaths than expected

42 had more deaths

7 did not report

THE FULL LIST

Hospitals ranked by mortality ratio, 2004-2007 average (all cases including patients in palliative care)

1	Regina General Hospital, Sask.	71
2	William Osler Health Centre Peel
	Memorial Hospital, Ont.	75
3	Royal University Hosp., Sask.	78
4	St. Mary's General Hosp., Ont.	81
5	St. Paul's Hospital, Sask.	81
6	Thunder Bay Regional Health
	Sciences Centre, Ont.	81
7	Wm. Osler Health Centre, Ont.	82
8	The Moncton Hospital, N.B.	83
9	Foothills Medical Centre, Alta.	84
10	University Health Network, Ont.	85
11	Hamilton Health Sciences
	Corp. Hamilton Division, Ont.	86
12	St. Joseph's Healthcare
	Hamilton, Ont.	86
13	Vancouver General Hosp., B.C.	86
14	Kelowna General Hosp., B.C.	87
15	Peter Lougheed Centre, Alta.	88
16	Saint John Rgnl. Hosp., N.B.	88
17	Joseph Brant Mem. Hosp., Ont.	89
18	Victoria General and Royal Jubilee Hospitals, B.C.	89
19	William Osler Health Centre
	Etobicoke General Hospital, Ont.	89
20	Univ. of Alberta Hospital, Alta.	91
21	York Central Hospital, Ont.	91
22	Lethbridge Rgnl. Hospital, Alta.	93
23	North York General Hospital,
	General Site, Ont.	93
24	Hamilton Health Sciences Corp.
	McMaster Division, Ont.	94
25	Rockyview General Hospital, Alta.	94
26	Royal Alexandra Hospital, Alta.	94
27	Quinte Healthcare Corp., Ont.	95
28	St. Michael's Hospital, Ont.	95
29	Hamilton Health
	Sciences Corp., Ont.	96
30	Brantford Community Healthcare
	System, General Site, Ont.	97
31	Dr. Everett Chalmers Regional
	Hospital, N.B.	97
32	Guelph General Hospital, Ont.	97
33	Royal Columbian Hospital, B.C.	97
34	Southlake Regional
	Health Centre, Ont.	97
35	Lions Gate Hospital, B.C.	98
36	Pasqua Hospital, Sask.	98
37	The Ottawa Hospital
	Civic Campus, Ont.	98
38	Toronto East General Hosp., Ont.	98
39	Credit Valley Hospital, Ont.	99
40	Hotel Dieu Grace Hospital, Ont.	99
41	Langley Memorial Hospital, B.C.	99
42	Nanaimo Regional
	General Hospital, B.C.	99
43	Sault Area Hospitals, Ont.	99
44	Lakeridge Hospital Corp., Ont.	100
45	The Ottawa Hospital, Ont.	100
46	Rouge Valley Health Syst., Ont.	101
47	The Ottawa Hospital General
	Campus, Ont.	101
48	Lakeridge Health
	Oshawa Site, Ont.	102
49	Matsqui Sumas Abbotsford
	General Hospital, B.C.	102
50	Sunnybrook Health Sciences
	Centre, Ont.	102
51	Trillium Health Centre
	Mississauga, Ont.	102
52	Grey Bruce Health Services, Ont.	103
53	Queensway Carleton Hosp., Ont.	103
54	Peace Arch District Hospital, B.C.	104
55	Rouge Valley Health System
	Centenary Site, Ont.	104
56	Royal Inland Hospital, B.C.	104
57	Royal Victoria Hospital of
	Barrie, Ont.	104
58	Halton Healthcare Services
	Corporation, Ont.	105
59	Providence Health Care, B.C.	105
60	St. Joseph's Health Centre,
	Toronto, Ont.	105
61	St. Paul's Hospital, B.C.	105
62	Chatham Kent Health
	Alliance, Ont.	106
63	Peterborough Regional Health
	Centre, Ont.	106
64	St. Boniface General Hosp., Man.	106
65	Bluewater Health, Sarnia Site, Ont.	107
66	Sudbury Regional Hospital,
	Laurentian Site, Ont.	107
67	Halton Healthcare Services,
	Oakville Trafalgar Site, Ont.	109
68	Health Sciences Centre, Man.	109
69	Kingston General Hospital, Ont.	109
70	Queen Elizabeth II Health
	Sciences Centre, N.S.	109
71	Sarnia Bluewater Health, Ont.	109
72	Hamilton Health Sciences Corp.
	Henderson Division, Ont.	111
73	Mount Sinai Hospital, Ont.	111
74	St. John's Acute Care, Nfld./Lab.	111
75	Surrey Memorial Hospital, B.C.	112
76	London Health Sciences
	Centre, Ont.	114
77	Red Deer Regional Hospital
	Centre, Alta.	115
78	Burnaby Hospital, B.C.	116
79	The Scarborough Hospital,
	Grace Site, Ont.	116
80	Niagara Health System, Ont.	122
81	The Scarborough Hospital, Ont.	124
82	Niagara Health System,
	St. Catharines General Site, Ont.	129
83	The Scarborough Hospital,
	General Site, Ont.	129
84	Grand River Hospital,
	K. W. Health Centre, Ont.	130
85	Cape Breton Healthcare
	Complex, N.S.	137
86	Grey Nuns Community
	Hospital and Health Centre, Alta.	NR
87	Humber River Regional
	Hospital, Ont.	NR
88	Humber River Regional Hospital,
	Church Street Site, Ont.	NR
89	Humber River Regional
	Hospital, Finch Site, Ont.	NR
90	Misericordia Community
	Hospital, Alta.	NR
91	Windsor Regional Hospital, Ont.	NR
92	Windsor Regional Hospital,
	Metropolitan Campus, Ont.	NR