The Yield and Safety of Screening and Surveillance Colonoscopy in Elderly Patients (> 80 Years)
Background and Clinical Significance:
Colorectal cancer continues to carry a significant burden of morbidity and mortality into
the twenty-first century, despite the availability of multiple screening modalities. It is
estimated that approximately 150,000 new cases of CRC will be diagnosed and over 50,000
people will die of CRC in 2006 (1). Current options for screening for CRC include fecal
occult-blood testing, flexible sigmoidoscopy, double contrast barium enema, and colonoscopy.
While no randomized controlled trials have shown a mortality benefit with screening
colonoscopy, it has become the preferred method of both screening and surveillance of polyps
because of the ability to visualize the entire colon in addition to having the ability to
remove polyps. This procedure does associated risks including perforation and bleeding
after polypectomy. Other major complications have also been reported, including MI and CVA.
Although guidelines exist for colorectal cancer screening and surveillance of polyps, they
do not define the upper age limit to which these practices should be carried out (2,3). No
clear data is available on the effect of this procedure on life expectancy after the age of
80. Also, the country has limited resources and screening colonoscopy may be offered to
those who are most likely to benefit. Based on these facts, there is a real need to quantify
the prevalence of colon neoplasia in this age group to guide primary care physicians as well
as gastroenterologists in offering screening and surveillance colonoscopy to this age group.
Like many medical decisions, cancer screening requires weighing quantitative information,
such as risk of cancer death and likelihood of beneficial and adverse screening outcomes, as
well as qualitative factors, such as individual patients' values and preferences. In fact,
patients with life expectancies of less than 5 years are unlikely to derive any survival
benefit from cancer screening. There is also potential for harm from screening procedures.
Some of the greatest harms of screening occur by detecting cancers that would never have
become clinically significant. This becomes more likely as life expectancy decreases.
As the population ages and life expectancy continues to increase, more elderly patients will
be referred for colonoscopy. What is the utility of performing colonoscopy in these
asymptomatic patients, and is it safe? Few studies have examined this question adequately.
What follows is a brief review of published data in regards to yield and safety of
colonoscopy in the elderly.
Lin et al examined the prevalence of neoplasia in 1244 screening colonoscopies - 63 were
patients older than 80. They found that although the prevalence of neoplasia increases with
age (28.6% in patients older than 80), the gain in expected life expectancy after
intervention is limited (4).
Cooper et al came to a somewhat different conclusion (7). They analyzed 1.8 million
Medicare patient colonoscopies in 1999 using ICD-9 codes for rates of polyp detection
(pathology was not examined, so it is not clear what percentage of these patients had
hyperplastic polyps, adenomas, advanced adenomas, or cancer). They found that the rate of
polyp detection decreased with age but was still high (ranging from 15.2-31.3% in patients
older than 80). Cooper had previously shown that the incidence of colorectal cancer
increased with age in an analysis of all new cases of colon cancer in Medicare patients in
19878. Similarly, the National Polyp Study has shown an increase in the incidence of high
grade dysplasia in patients older than 60 (9).
Finally, the VA Cooperative Study Group No. 380 did a prospective cross sectional study of
3121 asymptomatic patients between 1994-1997 to find the prevalence of advanced neoplasia
and associated risk factors. 329 patients were found to have advanced neoplasia as defined
by an adenoma greater than 1 cm, villous histology, presence of high-grade dysplasia, or
cancer. Associations were found between advanced neoplasia and family history of CRC,
smoking, moderate to heavy alcohol consumption, fiber intake, and use of daily NSAIDS or
Two of the above studies discussed safety of colonoscopy in elderly patients. Duncan found
8 major complications (0.6%) including 3 bleeds and 1 perforation in a series of 1199
colonoscopies6. Sardinha also reported a low rate of major bleeding (0.2%) and no
perforation in 428 colonoscopies; this compared favorably with two other studies on elderly
patients.10 Gatto et al took a random sample of Medicare patients who underwent a
colonoscopy between 1991 and 1998. 39,286 colonoscopies were identified by CPT-4 codes;
perforations within 7 days of the procedure were identified by ICD-9 codes. The average age
was 74, and 21 % of patients were older than 80. The overall incidence of perforation was
0.19%; the authors found that the rate of perforation increased with age and the number of
In addition to perforation and bleeding, MI and CVA have also been reported after
colonoscopy. Cappell studied patients who were already at higher risk for an MI or CVA. He
looked at 100 patients who underwent colonoscopy within 30 days after MI and compared them
to 100 control patients without MI or unstable angina in the preceding 6 months. He found
that while there was a higher rate of minor complications (transient asymptomatic
hypotension or bradycardia) in the study group (in which the patients were sicker overall),
there was only one major complication after colonoscopy which was probably not due to the
Importance of Current Research to Veterans:
The prevalence of CRC is approximately 5% (1,3). Although over 80% of cases are sporadic
(the remaining result from inflammatory bowel disease and hereditary colon cancer
syndromes), many veterans have one or more risk factors for CRC including advancing age,
smoking, heavy alcohol intake, high fat low fiber diet, sedentary lifestyle, and obesity.
As the veteran population ages, preventative services will continue to be an important part
of their health care. As the US population is aging, so too is the US veteran population.
This is further complicated by the fact that a recent study that examined the self-rated
health and functioning of a national sample of veterans aged 65 or older reported that over
one-half of elderly veterans report difficulty in functioning and rated their health status
as fair or poor (14). This group of veterans may also live alone, may not have easy access
to transportation to and from a health facility and may find it physically challenging to
prepare their colon with cleansing agents for a good exam during a colonoscopy. Age and
comorbidity also affect the survival rates after resection of colorectal cancer. A study by
Ko et al used nationwide data from the healthcare cost and utilization program and
calculated mortality among 22,000 resections for colon cancer15. They found that besides the
volume of surgery, two other factors that affect post colon cancer mortality are age and
comorbid disease (i.e. cardiovascular, pulmonary and liver diseases) (p<0.05). Another study
of 80 patients (16), 80 years and older, showed a postoperative mortality rate of 8% and 5yr
survival of only 23%. Only 13 patients lived longer than 5 years. Many elderly people may
need a coronary revascularization procedure before planned surgery, if a colon cancer is
discovered. Older age has been associated with adverse outcomes in patients undergoing
percutaneous coronary intervention. A study by the National Heart, Lung and Blood Institute
included 4620 PCI treated patients and reported adjusted relative risks in elderly (>80
years) compared to younger (<65 years) patients to be higher for inhospital related death
(3.64 versus 1.0) as well as myocardial infarction (2.57 versus 1.0) (17).
Based on the above data, it is useful and necessary to know the potential risks and benefits
of this invasive procedure in otherwise asymptomatic elderly patients.
Study Design: Computerized medical records and endoscopy reports will be searched to
identify patients older than 80 years old who underwent colonoscopy indicated for average
risk screening for colon cancer or surveillance of polyps. A control group of patients
between the ages of 50 and 79 will also be identified who underwent colonoscopy indicated
for average risk screening for colon cancer or surveillance of polyps. The ratio of
controls to study patients will be 2 to 1. Records will be searched from 1997 to July 31,
2006. Data collected will include age at time of colonoscopy, sex, body mass index, a
personal history of CAD, DM, CVA, or PVD, endoscopic and histologic findings of colonoscopy,
evidence of complete examination, all complications during colonoscopy, evidence of
complications after colonoscopy including perforation, bleeding, MI, and CVA. The use of
aspirin, NSAIDS, calcium, and vitamin D will be recorded. A comprehensive questionnaire that
will include information on diet, exercise, family history, smoking, alcohol consumption,
use of aspirin, NSAIDS, calcium, vitamin D, and statins will be administered by the
endoscopist prior to colonoscopy.
Risk and Benefit to the Study Participant:
The participants will only be active in the prospective portion of the study. Their
participation will be limited to completion of a written survey. This will be given after
verbal informed consent. There will be no risk to the study participant. Benefits of this
study have been previously addressed in detail in the section titled " Importance of Current
Research to Veterans".
Patient demographics, endoscopy and pathology findings, procedure complications,
comorbidities and data regarding various risk factors such as smoking, alcohol intake, lack
of dietary fruits and vegetables, lack of exercise, family history, prior history of colon
cancer, ASA, NSAID and statin use will be collected. A questionnaire for the factors
outlined above will be administered by an endoscopist. The data will be transferred to an
Excel spreadsheet by one of the study investigators.
Statistical Power and Data Analysis:
Using the SPSS software, the prevalence of colon polyps and colon cancer between the group
age >80 years and < 80 years will be compared. Significant polyps will be defined as polyp >
1 cm in size, polyps with villous histology, three or more polyps and polyps with dysplasia.
Hyperplastic polyps will be considered non-neoplastic. Smokers will be defined as people
who smoked at least 1 pack per day for 10 years and who did not quit smoking in the past 10
years. Associations between the clinical and demographic variables will be examined using
Spearman's correlation coefficients. Univariate analysis will be performed Chi-square or
fisher's exact test for categorical variables and Wilcoxon sum test for continuous
variables. A multivariate logistic regression will be performed to control for the
covariates that may affect the prevalence of colonic neoplasia. All tests will be
two-tailed. A p value of < 0.05 will be considered significant.
Observational Model: Case Control, Time Perspective: Retrospective
Proportion of elderly patients (>80 yrs) with colorectal neoplasia
Ajay Bansal, MD
Kansas city VA Medical Center
United States: Federal Government
|Kansascity VA Medical center||Kansas city, Missouri 64128|