The following letter was emailed to the Times Colonist late on Friday afternoon.

May 9, 2003

Times Colonist

Dear Editor:

On May 9th you published Health Planning Minister Sindi Hawkins’ letter in which she claimed that the increase in infant mortality is not statistically significant in two categories. Hawkins and her government should spend more time identifying and dealing with preventable deaths and less time trying to dismiss the 20% increase as insignificant. Her government removed all infant mortality except aboriginal from the performance measures in all government service plans.

In fact the increase in infant mortality from 3.70 deaths per 1,000 in 2000 to 4.44 deaths per 1,000 in 2002 is statistically significant at the 95% level if one constructs a confidence level around the 2002 number and asks the question whether it is different from 3.70. The Ministry of Health Planning uses a more difficult test that puts a confidence interval around both years and then examines whether the two confidence intervals overlap. The increase is statistically significant at the 75% level using that more stringent test.

An increase of one preventable death would be one too much, but a 20% increase that is statistically significant is very troubling. Whether the increase is due to welfare payments that have been cut, child protection investigations that are down by 20%, health care that is being reorganized or any other factor that can be modified, every effort should be made to identify preventable deaths and take corrective action.

(For background on performing the significance tests see: Statistical Notes for Health Planners. No. 13. Changes in Infant Mortality and Related Rates by Health Service Area: 1969-73 to 1974-77. Foster, J. E.; Kleinman, J. C. May 13, 1981. 32 pp. (PHS) 81-1237 which is available at http://www.cdc.gov/nchs/data/statnthp/statnthp13.pdf.)

---David Schreck

---

In response to my letter of May 8th, Dr. Kendall sent an email saying:

"We may be using slightly different methods to calculate these rates - the ministry uses that adopted by the BC Vital Statistics Agency (see p 127 of their Annual Report for details)- but the results of our calculations agree with yours."

"The confidence intervals we calculate for the 2002 IMR (4.44) are:- upper limit 5.09, lower limit 3.79"

"The confidence intervals around the 2000 IMR (3.70) are:- upper limit 4.30, lower limit 3.11."

"In other words the 95% confidence intervals around the 2000 and the 2002 rates overlap, and statistically this means there is no significant difference at the 95% confidence level between the rates in these two years."

My reply is copied below:

Dr. Kendall:

Thank you for your note. That is one way to quibble about the statistics but I would hope that as Provincial Health Officer you would follow the long standing tradition of advocacy in public health. A 20% increase over 2 years is a concern. The 4.44 rate is significantly different from 3.70, although by adopting a test of the difference in proportions you can deny that the 27 additional deaths are something to worry about. One way or the other, every death should be examined to determine whether it could have been prevented. I am disappointed that the impression left in the CKNW story was that there is nothing to be concerned about. Welfare rates have been cut, the Ministry of Children and Family Development is having its budget cut by over $300 million, investigations of child abuse have decreased by 20% (see http://www.StrategicThoughts.com/record2003/protect.html) and surgical waiting lists have increased by 20%. I am not saying that the increase in infant mortality is the direct result of those changes, but you know as well as I that infant mortality is a key indicator of the consequences of such regressive policies. That is why you have a particularly important role in investigating each death rather than making public statements that there is no problem.

I look forward to seeing your detailed report. Based on the work you did on your child mortality report published in May 2001, I have every confidence that you will uphold the best traditions of advocacy in public health.

---

Copied below is the second letter I faxed to Dr. Kendall in response to the report on the CKNW website at http://www.cknw.com/news/localnews_detail.cfm?item=23547

May 8, 2003

Dr. Perry Kendall
Provincial Health Officer
Ministry of Health Planning
Main Floor, 1520 Blanshard St                                 by fax to (250) 952-1362 (& emailed)
Victoria, British Columbia V8W 3C8

Dear Dr. Kendall:

Thank you for your email last evening where you indicated that you would get back to me with your statistical analysis of BC's recent increase in infant mortality together with International Classification of Disease 10 categorization of recent deaths. I hope the ICD 10 information can be made available by year for the past several years.

I calculated a 95% confidence interval for BC's 2002 infant mortality rate. That confidence interval has an upper limit of 5.09 and a lower limit of 3.79. In other words, the year 2002 infant mortality rate of 4.44 is significantly different than the year 2000 rate of 3.70 at the 95% confidence level.

The methodology I used is the same as that described on page 5 of Statistical Notes for Health Planners. No. 13. Changes in Infant Mortality and Related Rates by Health Service Area: 1969-73 to 1974-77. Foster, J. E.; Kleinman, J. C. May 13, 1981. 32 pp. (PHS) 81-1237 which is available at http://www.cdc.gov/nchs/data/statnthp/statnthp13.pdf.

Unless you have some other statistical test that would meet the standards of the Center for Disease Control, I encourage you to correct the false impression that was left in the CKNW broadcast which claimed that you had said that the increase is not yet statistically significant. More importantly, I hope you will work with Jane Morley to identify the cases of infant deaths that were preventable in the same manner that you did in your May 2001 publication on child mortality.

Sincerely,

David D. Schreck

---

Copied below is the letter I have faxed to Dr. Kendall in response to the report on the CKNW website at http://www.cknw.com/news/localnews_detail.cfm?item=23547

May 7, 2003

Dr. Perry Kendall
Provincial Health Officer
Ministry of Health Planning
Main Floor, 1520 Blanshard St                   by fax only to (250) 952-1362
Victoria, British Columbia V8W 3C8

Dear Dr. Kendall:

On May 7 CKNW attributed a statement to you saying that while infant mortality has increased in the last two years “the increase is not yet statistically significant.” Could you please say what would be statistically significant, and what statistical test of significance you applied to the numbers?

In your excellent report titled "Health Status of Children and Youth in Care in British Columbia What do the Mortality Data Show?", you wrote "Based on child fatality reviews, the Children's Commission provides recommendations for preventing further child deaths - a major step forward since the years prior to the Commission's establishment in 1996. With continued efforts and a planned, systematic approach, we can expect to see child and youth deaths become even more rare in the future."

In the absence of a Children’s Commission, will there be a child fatality review and recommendations for preventing further deaths? Can you separate the various causes of infant death over the last two years as you did in that report and identify how many of the deaths were preventable?

Sincerely,

David D. Schreck

---

April 29, 2003

Infant Mortality Up 10% Per Year
in Campbell's New Era

Even though the Campbell government has eliminated infant mortality as a measure of government performance, BC's Vital Statistics continues to collect the data. The Campbell government should be concerned that infant mortality has risen by 10% each year since it took power. Is that misfortune, or is that a measure of failed policies?

It will be several weeks before Vital Statistics publishes data for the full calendar year 2002 but a review of the quarterly reports indicated a surprising increase so I requested an early copy of the infant mortality data for 2002. The following table shows both the total infant deaths and the rate per thousand live births for 1998 through 2002. The rate increased by 10% in 2001, and by 10% again in 2002! In absolute terms, there were 27 more infant deaths in 2002 than in the low year of 2000 even though there were 623 fewer births in 2002.

Deaths of Infants under 1 Year of Age

	infant deaths	live births	Infant Death Rate (per 1,000 livebirths)
1998	174	42,862	4.06
1999	158	41,748	3.78
2000	150	40,493	3.70
2001	162	40,391	4.01
2002	177	39,870	4.44

The graph shown to the right is from page 51 of the 2001/02 Annual Report for the Ministry of Health Services. It shows that the infant mortality rate declined throughout the period of NDP government. It reached record lows for both the general population and for the aboriginal population in 2000. It doesn't show the statistics since 2000 when the downward trend reversed. The Campbell government has said that it will use infant mortality for the aboriginal population as an important measure in its service plan for the Ministry of Health Services but it will not use infant mortality for the overall population as a measure of performance for any aspect of government. The public may be shocked about the government's failure to accept that important measure, and it very likely will be concerned over the rising number of deaths.

The Campbell government eliminated the positions of the Children's Commission and the Child, Youth and Family Advocate. Those independent officers have been replaced with the "children and youth officer" who reports to the Attorney General. The appointment of Jane Morley, QC, to that position takes effect on May 1st. If the former independent positions still existed, it is very likely that each of the infant deaths that have occurred would be investigated and a report would be submitted to the legislature with respect to any weaknesses or failures in government policy that contributed to those deaths. Eyes will be on Morley to see whether she follows that precedent and holds government accountable.

---

"The release of this report is consistent with our government's New Era commitments to increased accountability and a stronger focus on health planning and health quality. Our emphasis on performance measurement and accountability is reflected in three-year ministry service plans, ministry annual reports, and performance agreements with health authorities."

"The infant mortality rate is a long established measure of child health that also indicates the well-being of a society. It reflects not only the level of mortality, but also the health status and health care of population, the effectiveness of preventive care and the attention paid to maternal and child health, as well as broader social factors such as maternal education, smoking, and relative deprivation."