NAACCReview

Archives for May 2015

Benign prostatic hyperplasia mortality in California

frankboscoeFrancis P. Boscoe, Ph.D, Research Scientist, New York State Cancer Registry (NAACCR at-large Board Member)

Why are so many men in California dying of hyperplasia of the prostate? Many people have been asking this question lately in response to an article I wrote identifying the most distinctive causes of death in every state, which has been receiving a lot of media coverage and circulation in the blogosphere.

Hyperplasia of the prostate is ICD-10 code N40, and goes by the synonyms of enlarged prostate, nodular prostate, polyp of prostate, benign prostatic hypertrophy, adenofibromatous hypertrophy of prostate, and, most familiarly, BPH. It does not include either benign or malignant prostate neoplasms, which have ICD-10 codes D29.1 and C61, respectively.

There were 937 deaths due to BPH in California in the period 2001-2010, for an age-adjusted rate of 8 per million men, compared with 4 per million nationally. The real question is perhaps not why California is double the national average, but why even the national average is as high as it is. Should anyone at all be dying from a condition that includes the word “benign”?

It turns out that BPH behaves like benign brain tumors that are familiar to many readers of this blog: the condition itself is benign, but can become lethal once the growth reaches a point that it begins to impede the function of nearby tissues or organs. In the case of BPH, the prostate can eventually become large enough that it partially or completely blocks the urethra, leading to inability to urinate, urinary tract infections, bladder and kidney damage, and if left completely untreated, ultimately to death.

Given that BPH is a treatable condition, one could reasonably argue that a death rate of 8 per million or even 1 per million is unacceptably high. While it is true that death rates have dropped dramatically since the 1970s – by 95% according to one English study – they have leveled off in recent years (in California, the age-adjusted mortality rate has varied between 6 and 9 per million men each year from 1999 to 2013). Also, given the number of people I’ve spoken with over the past week who thought that the appearance of BPH on the map must represent some kind of error – even some who are medically knowledgeable – I think calling attention to this rare but preventable cause of death has been useful.

The NIH has a helpful fact sheet on BPH here.


Click here to view original article (The abstract below is from the Centers for Disease Control and Prevention article ‘The Most Distinctive Causes of Death by State, 2001-2010’)


Abstract

The most distinctive cause of death (defined as the location quotient) for each state and the District of Columbia, 2001–2010. The map shows the cause of death from the International Classification of Diseases, 10th Revision (ICD-10), List of 113 Selected Causes of Death with the highest age-adjusted mortality rate ratio in each state. The causes are listed in the legend in the order of disease classification in ICD-10. This map highlights nonstandard cause-of-death certification practices within and between states that can potentially be addressed through education and training.

14_0395_large_01

Click image to enlarge

Background: Maps of the most distinctive or characteristic value of some variable at the state or country level became popular on social media in 2014. Among the most widely shared examples have been maps of state-level birth name preferences, music-listening preferences, and mortality from among the top 10 causes of death (1). This form of data presentation has a long history in economic geography, where the mapped values are known as location quotients (2). We use the International Classification of Diseases, 10th Revision (ICD-10), List of 113 Selected Causes of Death file published by the National Center for Health Statistics (3) to present a more nuanced view of mortality variation within the United States than what can be seen by using only the 10 most common causes of death.

Methods: Counts for each cause of death included on the ICD-10 List of 113 Selected Causes of Death along with population sizes were obtained for each of the 50 states and the District of Columbia for 2001 through 2010 from the Underlying Cause of Death file accessible through the Centers for Disease Control and Prevention (CDC) WONDER (Wide-ranging Online Data for Epidemiologic Research) website (4). We also included subcauses of death contained in this file, such as specific types of cancer, which brought the total number of causes of death to 136. The standardized mortality rate ratio (ie, the ratio of the age-adjusted state-specific death rate for each cause of death relative to the national age-adjusted death rate for each cause of death, equivalent to a location quotient) was then calculated, and the maximum ratio for each state was mapped. That is, we mapped

cdc-frank-boscoe

where Maxj is the age-adjusted mortality rate for each state i and SMRij is the age-adjusted mortality rate for the United States for each cause of death j. Causes of death with fewer than 10 counts at the state level were suppressed and therefore not available for this analysis.

The map was produced in SAS software version 9.3 (SAS Institute, Inc) by using a single program that imported the output from CDC WONDER, calculated the mortality rate ratios, and generated the map using PROC MAPIMPORT and PROC GMAP. The program code is available from the authors. Minor cosmetic enhancements were made to the map using Adobe Illustrator (Adobe, Inc). Both colors and numeric labels were used on the map to facilitate black-and-white printing.

Main Findings: The resulting map depicts a variety of distinctive causes of death based on a wide range of number of deaths, from 15,000 deaths from HIV in Florida to 679 deaths from tuberculosis in Texas to 22 deaths from syphilis in Louisiana. The largest number of deaths mapped were the 37,292 deaths in Michigan from “atherosclerotic cardiovascular disease, so described”; the fewest, the 11 deaths in Montana from “acute and rapidly progressive nephritic and nephrotic syndrome.” The state-specific percentage of total deaths mapped ranged from 1.8% (Delaware; atherosclerotic cardiovascular disease, so described) to 0.0005% (Illinois, other disorders of kidney).

Some of the findings make intuitive sense (influenza in some northern states, pneumoconioses in coal-mining states, air and water accidents in Alaska and Idaho), while the explanations for others are less immediately apparent (septicemia in New Jersey, deaths by legal intervention in 3 Western states). The highly variable use of codes beginning with “other” between states is also apparent. For example, Oklahoma accounted for 24% of the deaths attributable to “other acute ischemic heart diseases” in the country despite having only slightly more than 1% of the population, resulting in a standardized mortality rate ratio of 19.4 for this cause of death, the highest on the map. The highest standardized mortality rate ratio after Oklahoma was 12.4 for pneumoconioses in West Virginia.

A limitation of this map is that it depicts only 1 distinctive cause of death for each state. All of these were significantly higher than the national rate, but there were many others also significantly higher than the national rate that were not mapped. The map is also predisposed to showing rare causes of death — for 22 of the states, the total number of deaths mapped was under 100. Using broader cause-of-death categories or requiring a higher threshold for the number of deaths would result in a different map. These limitations are characteristic of maps generally and are why these maps are best regarded as snapshots and not comprehensive statistical summaries (5).

Action: This map has been a robust conversation starter among those who have seen it before publication, generating hypotheses and inviting further exploration of the underlying data set, something that an equivalent tabular representation does not accomplish as well. Although chronic disease prevention efforts should continue to emphasize the most common conditions, an outlier map such as this one should also be of interest to public health professionals, particularly insofar as it highlights nonstandard cause-of-death certification practices within and between states that can potentially be addressed through education and training. This is especially true considering that most death certificates are completed by community physicians who receive little or no formal training in this area. For example, a study found that nearly half of the death certificates certified by physicians in a suburban Florida county contained major errors, often reflecting confusion between the underlying cause of death and the terminal mechanism of death (6). It would not take many systematic miscodes involving an unusual cause of death for it to appear on this type of map.


The opinions expressed in this article are those of the authors and may not represent the official positions of NAACCR.

As I lay dying

Laurie Becklund: Treat me like a statistic and save my life

Laurie Becklund: ‘Treat me like a statistic and save my life.’

I am dying, literally, at my home in Hollywood, of metastatic breast cancer, the only kind of breast cancer that kills. For six years I’ve known I was going to die. I just didn’t know when.

Then, a couple of weeks before Christmas, a new, deadly diagnosis gave me a deadline. No doctor would promise me I’d make it to 2015.

Promise me, I told my friends and family, that you’ll never say that I died after “fighting a courageous battle with breast cancer.” This tired, trite line dishonors the dead and the dying by suggesting that we, the victims, are responsible for our deaths or that the fight we were in was ever fair.

Promise me you’ll never wear a pink ribbon in my name or drop a dollar into a bucket that goes to breast cancer “awareness” for “early detection for a cure,” the mantra of fund-raising juggernaut Susan G. Komen, which has propagated a distorted message about breast cancer and how to “cure” it.

I’m proof that early detection doesn’t cure cancer. I had more than 20 mammograms, and none of them caught my disease. In fact, we now have significant studies showing that routine mammogram screening, which may result in misdiagnoses, unnecessary treatment and radiation overexposure, can harm more people than it helps.

In 1996, during a self-exam, I found a peanut-sized lump in one breast that turned out to be stage one breast cancer. I had the “best,” most common, kind of breast cancer, found it early, got a lumpectomy and short dose of radiation. Five years out, my doctor told me there was little chance of recurrence and said, “Have a great life!”

You can imagine my shock when, 13 years after my initial diagnosis, I was in gridlock on the Harbor Freeway and got a call from my doctor with the results of a PET scan ordered after routine blood labs. “Maybe you should pull over,” he said.

Half an hour later, in an elementary school parking lot, I learned the scans revealed stage four breast cancer in my bones, liver, lungs and brain: a death sentence with an average life expectancy of three years.

I demanded the truth, always, from my doctors. I was a reporter who needed facts to plan whatever life I had left. I would not live in denial. But I was too scared, too private to tell anyone except my husband, my daughter and three friends. My very cells suddenly became my most intimate secrets.


Read Full Article (Excerpt of Article from a Op-Ed piece published in the LA Times by Laurie Becklund)


Comments by:
John W. Morgan, DrPH, CPH, Professor of Epidemiology, Loma Linda University School of Public Health Epidemiologist, Regions 4, 5, 7 & 10 of the Cancer Registry of Greater California (NAACCR Committee Member)

With great sadness, I read the autobiographic account of Laurie Becklund’s experience with breast cancer, including the heartbreaking conclusion. My sadness stems from the tragedy of her loss and because her story neglects to acknowledge successes in the “war on cancer” that are too numerous to list here. Since mammographic screening for breast cancer became available in the U.S. during the middle 1980s, age-adjusted risk of death for this disease has shown a steady decline. Unfortunately, this success, which has extended lives for tens of thousands of women, cannot guarantee equal benefits for all. Using population-based cancer registry data, researchers, cancer registrars and numerous other health professionals have dedicated their lives to cancer control and prevention research that has improved early detection, targeted therapies and quality of life among patients diagnosed with cancer.

During my youth, I have vivid memories of a classmate diagnosed with, what I know today was, acute lymphocytic leukemia (ALL). Like virtually all children with ALL at that time, she quickly succumbed to the ravages of the cancer. This was my first encounter with the death of a child that I knew and I recall the fear that affected every child in our school as we learned that there was no cure for ALL, the most common childhood cancer.

Thanks to discoveries made in cancer research, the story of childhood cancer has a different ending for most afflicted children today. Currently, more than 90% of children diagnosed with ALL achieve permanent remissions from their cancer, with most having normal life expectancies. Whether children afflicted with cancer, their doctors or researchers receive credit for these victories is unimportant to me. What is of supreme importance is the understanding that each victory in the war on cancer is personal and each is dependent upon research discoveries that require high quality cancer surveillance data and funding.

Today, an average of one in eight women will have a breast cancer diagnosis during her lifetime. While this number seems startling, research that improves detection of earlier stage cancer and treatment has elevated the 5-year relative survival for breast cancer to more than 89%. I do not know whether medical progress extended or improved the quality of life for Laurie Becklund, although I am consoled in the understanding that ongoing research will allow us to incrementally move closer to her hope for the cure that she so much deserved.


The opinions expressed in this article are those of the authors and may not represent the official positions of NAACCR.

Breast Cancer Screening Draft Recommendations

us-draftThe United States Preventive Services Task Force (USPSTF) has posted draft breast cancer screening recommendations for public comments (http://screeningforbreastcancer.org/) along with supporting information. The opportunity for public comment ends May 18th. The 2015 draft recommendations are similar to the 2009 recommendations:

  • The USPSTF recommends biennial screening mammography for women ages 50 to 74 years. (“B” rating meaning that USPSTF recommends the service and there is high certainty that the net benefit is moderate or there is moderate certainty that the net benefit is moderate to substantial).
  • The decision to start screening mammography in women prior to age 50 years should be an individual one. Women who place a higher value on the potential benefit than the potential harms may choose to begin biennial screening between the ages of 40 and 49 years. (“C” rating meaning that USPSTF recommends selectively offering or providing this service to individual patients based on professional judgment and patient preferences. There is at least moderate certainty that the net benefit is small.)

There was insufficient evidence to make a recommendation for women age 75 and over because women in this age group were not included in randomized trials of breast cancer screening. Similarly the USPSTF found insufficient evidence on the benefits and harms of tomosythesis (3-D digital mammography) as a primary screening method as well as insufficient evidence on the use of additional imaging technologies for women with dense breasts after a negative mammogram.

Breast density based on the American College of Radiology’s Breast Imaging Reporting and Data System (BI-RADS) is a well-established risk factor for developing breast cancer; however it has not been shown to increase the risk of dying from breast cancer. Higher breast density reduces both the sensitivity and specificity of mammography, decreasing the likelihood of detecting a tumor with mammography and increasing chances of false positive results. Currently 22 states have breast cancer density reporting laws which requires facilities that perform mammograms to notify women if they have dense breasts. The USPSTF draft recommendation concludes that more evidence is needed to understand how the frequency of screening might affect important health outcomes in women with dense breasts.

As with the 2009 recommendations, the draft statement has generated considerable discussion, and much around the role of the USPSTF recommendation in relation to the Affordable Care Act which requires coverage of preventive services without cost-sharing (e.g., copayment or deductible) under new health insurance plans or policies, if they have a USPSTF grade of A or B. Currently the Affordable Care Act utilizes the 2002 recommendation on breast cancer screening of the U.S. Preventive Services Task Force rather than the 2009 recommendations to ensure access to mammography for women ages 40-49. Click here to view USPSTF A and B Recommendations.


Learn more about these Breast Cancer Screening Draft Recommendations


croninkComments by:
Kathy Cronin, Ph.D, Deputy Associate Director, Surveillance Research Program (NAACCR Committee Member)

After many years of study, mammography recommendations still remain controversial. In developing the recommendations, the USPSTF considered both the harms and the benefits of mammography and assessed the evidence to determine whether there would be a net benefit to women who are screened. The benefits are measured in the reduction of mortality from breast cancer and reduction in morbidity associated with earlier detection and hopefully less treatment. The harms are more difficult to measure and assess. The most serious harm is associated with the detection and treatment of tumors that may not ever have been detected without screening which leads to unnecessary treatment, often referred to as overdiagnosis. Based on our SEER data we see that women with very early disease receive surgery, many also receive adjuvant treatment, and over 30% of women with in situ disease receive mastectomies even though it is not clear what proportion of these women would have progressed to invasive disease. The most common harm is false positive screening tests which lead to additional mammograms and possibly biopsies. The uncertainty during the period of follow up causes anxiety and stress for women in addition to the added costs to women and to the health care system. Although much work has been done to quantify the degree of harms and benefits of breast cancer screening, for women between 40-49 years of age, the USPSTF draft recommendation states that the decision to be screened should be an individual one in consultation with their doctor.

The USPSTF reached out to the Cancer Intervention and Surveillance Modeling (CISNET) group for the 2009 and 2015 updates for mammography screening to help quantify the harms and benefits associated with different screening intervals and various ages to start and stop screening. Modeling uses information available through clinical trials and observational data to gain a better understanding of disease progression before diagnosis. Understanding the natural history of breast cancer provides the basis for predicting how screening could disrupt the natural progression through earlier detection and treatment. Incidence rates from cancer registries along with population screening utilization gives critical insight into the impact of screening. Modeling using this information attempts to parse out the portion of the increase in incidence that is explained by screening versus changes in risk as well as the portion of the decline in mortality that resulted from screening versus treatment advances.

Available information, including modeling that heavily relies on registry data, helps clarify the benefits and harms associated with different ages and screening schedules. For example the draft recommendation of “C “ for women ages 40-49 does not reflect a lack of evidence of benefit, but rather a tipping of the scale in the direction of harms because of a smaller number of deaths avoided with screening in this age group. Similarly, the added harms of screening every year versus every other year appear to outweigh the additional benefits. The fact that the most serious harm of overdiagnosed cases is not observable, since it is indistinguishable from a women who benefited from early diagnosis and treatment, makes decisions related to screening particularly complex. Registries are uniquely positioned to contribute to this area of research, particularly through the linkages of tumor specimens with longer term outcomes recorded in the registry system. Also, research is underway to identify markers that distinguish between aggressive and indolent disease at diagnosis, thereby avoiding overtreatment associated with early detection and maximizing the utility of cancer screening. However, until we have the capability to differentiate those tumors that will progress from those that do not, determining when and how often to screen will continue to rely heavily on a woman’s preferences and individual circumstances. The information provided in the USPSTF draft statement provides women with scientific data on the benefits and harm associated with breast cancer screening so that they can make informed decisions with their doctors.


The opinions expressed in this article are those of the authors and may not represent the official positions of NAACCR.

Copyright © 2018 NAACCR, Inc. All Rights Reserved | naaccr-swoosh-only See NAACCR Partners and Sponsors