Definition.
An aneurysm is defined as a more than 50 increase in the diameter of an artery caused by a weakening of the arterial wall (Hallet 2008). Aneurysms of the aorta are the most common and may effect either the thoracic or abdominal portion of this artery. However, only 25 of aortic aneurysms affect the thoracic portion, striking both men and women equally. Abdominal aortic aneurysms (AAA) on the contrary are known to be three to six times more prevalent in men (Hallet 2008, Vardulaki 2000 cited in MASS 2002 1531). AAA are defined as an aortic diameter superior to three centimeters (Hallet 2008). Risk factors include smoking, hypertension, old age, family history, race and gender.
Statistical relevance
It is estimated that 0.5 to 3.2 of the population suffers from AAA (Hallet 2008) and that the disease is responsible for 2.1 of deaths in men over the age of 65 (MASS 2002). In the United States alone, this accounts for 9,000 deaths per year, 33,000 elective surgeries and 1400 to 2800 surgically related deaths (The Aneurysm Detection and Management Veterans Affairs Cooperative Study Group 2002).
Clinical aspects
Patients suffering from AAA rarely show clinical signs before the aneurysm ruptures (Lindholt 2005). When clinical signs are present, they tend to be discreet, depending on their size and growth rate. Though rare, rapidly growing aneurysms can be sensitive, especially before they rupture. In general, aneurysm patients that present clinical signs will complain of a constant pain in the lumbosacral region or will present an abnormal pulsating mass in the lower abdomen. In some cases, auscultation of the area will detect a systolic murmur (Hallet 2008). Unfortunately, the most commonly noted signs of AAA appear once the aneurysm has ruptured. Approximately half of these patients will die before arriving at a hospital and an additional 30 to 70 will die upon arrival (MASS 2002). This high mortality rate of 65 to 85 stresses the necessity for adequate screening of all patients at risk for AAA.
Diagnosis
AAA can be diagnosed through different imaging techniques. Ultrasonography is the most cost efficient, minimally invasive technique. Other possibilities include computed tomography or magnetic resonance angiography which can be useful in defining the aneurism s size and anatomy (Hallet 2008). In some cases, routine radiographs performed in search of other abnormalities may allow the practitioner to view a calcification of the affected arterial wall.
Treatment
A ruptured abdominal aortic aneurysm requires emergency surgery. Two surgical techniques exist. The first involves open laparotomy. This technique is associated with a 50 mortality rate. Endovascular stent grafting has a lower, though still considerable mortality rate of 20-30 (Hallet 2008). Mortality rates are high in both cases due to the precarious cardiovascular state these patients arrive in and to preexisting concomitant cardiovascular diseases. In cases where AAA is detected before rupture occurs, elective graft surgery may be proposed depending on the risk of rupture but also on the presence or not of contraindications for surgery. Indeed, because the elective surgery is associated with a 2-6 mortality rate, it is important to carefully select patients for this preventive surgery (MASS 2002). Conservative treatment consists of controlling risk factors such as hypertension.
Screening is an important measure in controlling AAA. Patients diagnosed with this ailment must be followed on a regular basis and informed of the possibility of elective surgery and of it s associated mortality rates. Ultrasonography is the technique of choice as it minimizes the patient s exposure to radiation. The following articles discuss different screening programs.
Immediate Repair Compared with Surveillance of Small Abdominal Aortic Aneurysms
The study clearly stated it s objective to compare the outcome of patients diagnosed with small AAA according to their treatment plan a) immediate surgical repair, or b) ultrasonographic or CT surveillance with surgery reserved for high risk aneurysms. Patients 50 to 79 years of age who had been diagnosed by CT with an AAA 4.0 to 5.4 cm in diameters within 12 weeks of the beginning of the study were included. A complete list of exclusion criteria was also provided in the article. The study was then correctly carried out as a RCT to avoid bias in treatment assignment, to facilitate blinding and to allow the use of the probability theory.
Patients were randomly assigned to one of two blocks (immediate repair or surveillance) by a computer generated random-number code. The study was composed of two equally sized groups of over 500 patients each. This allowed to statistically prove the results with a confidence level of 95. Randomization was stratified for each medical centre to ensure a proper balance of participants characteristics between groups.
The two groups were very similar at baseline. Considering the surgical nature of the treatment made it was impossible to conceal to which group patients had been allocated. To compensate for this, neither patients nor investigators had access to the outcome data during the study and the causes of all deaths were determined by a blinded outcomes committee. CT scans were also evaluated by a central laboratory. Though not perfectly blinded, all efforts were made to eliminate observer bias.
The status (dead or alive surgically repaired or not) of all 1136 patients was accounted for at the end of the study and the group related outcomes evaluated through intention-to-treat analysis. 92.6 of immediate-repair patients and 61.6 of the surveillance group patients had undergone surgery. It was unclear why the remaining 7.4 of immediate-repair patients had not undergone surgery. Data from surviving patients was collected through follow-up visits and final CT scans. Causes of death were determined by an outcomes committee.
Results were presented as a relative risk of death, either from AAA-related causes or all causes. No significant difference in outcome was noted between the immediate-repair group and the surveillance group.
A confidence level of 95 was reported for all results. The results are precise enough to encourage patients to follow a surveillance program rather than opting for immediate elective surgery for small aneurysms.
This study proved that immediate-repair was not associated with a higher survival rate but rather with more hospitalizations and therefore a higher cost than a surveillance program. Unfortunately, the majority of the patients included in this study were men. Though their is a higher prevalence of AAA among older men, the study would have been more interesting had it included more women, especially considering that women are at a higher risk for AAA rupture. The results of this study are therefore relevant for men but not necessarily for women.
Screening for abdominal aortic aneurysms single centre randomised controlled trial
The study was clearly focused on determining if AAA screening reduced mortality rates in Danish men (64 to 73 years old) from Viborg County. It was appropriately performed as a RCT, comparing a screening group to a control group. The means by which 12,639 men were randomly assigned to either the screening group (6333) or the control population (6306) are not described. Though the groups were balanced in numbers, little data is available making it impossible to judge if the two groups were statistically similar at baseline.
The patients were not blinded and were informed of the risk factors for AAA, possibly causing a behavior bias. All outcome data was analyzed by two blinded surgeons, thus eliminating observer bias. All 12,639 men who entered the trial were accounted for (dead or alive treated or not). No patients were transfered from their original group even though 1481 men from the screening group did not attend screening. Analysis of all outcome data was carried out on an intention-to-treat principal.
Patients in both groups were followed up for 0 to 69 months. Though the two groups were apparently similar for duration of follow-up, this still represents a large variation in data collection. Data on deaths and surgical procedures were collected by the same means in both groups.
The results are presented as relative risks with a confidence interval of 95 and appropriate p values. They are very precise with very low p values and acceptably small intervals. The results prove that screening significantly increases the number of men to undergo elective surgery and decreases the number of ruptured AAA as well as the number of AAA-related deaths. The benefits of screening on outcome are as clear at the lower confidence level as at the higher level.
All outcomes (relative risk of outcome, elective or emergency surgery and AAA rupture) were properly considered between the two groups. The study excluded women and gave very little information on it s population, thus making it difficult for readers to compare the study groups to any local populations. However, the discovery that only 352 men must be screened to save one life makes the results of this study particularly interesting, even considering the study s shortcomings.
The Multicentre Aneurysm Screening Study into the effect of abdominal aortic aneurysm screening on mortality in men a randomised controlled trial.
The study s clearly stated objectives were to evaluate the mortality rate (AAA-related and not) and the quality of life of men, aged 65 to 74 years living in the UK, screened for AAA in comparison with a control population. The study was therefore appropriately designed as a RCT with patients being randomly assigned to either a screening group or a control group. Patients were randomly allocated to one of the two groups through computer generated pseudorandom numbers. Randomization was then balanced at baseline for screening centre, age and social deprivation scores.
Attempts were made to carry out a blinded study. For instance, a data monitoring committee was responsible for evaluating death related information and the sponsors of the trial played no role in the study.
Results were not provided to the trial coordinators until all data had been collected. 80 of the men invited for screening were evaluated. Final outcome (dead or alive) was available for 99 of the men initially randomized.
Clinical follow-up was available for 81 of patients in whom screening had detected an aneurysm. Data was evaluated following a pre-specified plan. Death certificates were obtained through the Office of National Statistics and analyzed by a data monitoring committee. Surgical data was obtained through the medical centers and subsamples of each group completed identical tests to evaluate their quality of life following screening.
Before the trail began, it was determined that 66,000 men needed to be randomized to detect a 30 decrease in mortality with an 80 power of detection and a 5 significance level. This was based on approximately 115 ruptured AAA deaths in the control group. Over 70,000 men were therefore randomized which allowed to observe 91 deaths from ruptured AAA in the control population.
Results were presented as hazard ratios and corresponding absolute risks. AAA-related deaths were reduced in the screening group when compared to the control population (hazard ratio of 0.58). No significant change in mental health was noted in the men, regardless of their group and screening results. Outcome was evaluated with a 95 confidence level, a tight interval and a low p value of 0.0002. Mental health results were also subjected to low p values.
The financial aspect of screening was not discussed in this study. Though a reference was made to future publication of these results, a comment on the cost of screening would have been interesting.
Population based randomised controlled trial on impact of screening on mortality from aaa
The study aimed to discover if a screening program for AAA would reduce AAA-related mortality rates in Perth men aged 65 to 79 years. It was appropriately carried out as a RCT, comparing a study group to a control group. Participants were allocated to two equal sized groups balanced for age. No description of the randomization process is given making it impossible to evaluate if the process was appropriate or biased.
All attempts were made to eliminate observer bias. Australia s unified database system allowed for all data pertaining to the participants surgical procedures andor deaths to be accounted for at the end. Data was submitted to an intention-to-treat analysis. A blinded researcher evaluated death certificates for cause of death. When AAA was involved, a second blinded analysis was completed for confirmation.
All outcome data was collected through Australia s unified database system. Both groups suffered equal losses between randomization and screening. Low participation rates may have biased the results in contrast to other studies with stricter inclusion criteria and higher participation rates. It was estimated that a study involving a control group of 20,000 men experiencing about 55 AAA-related deaths would allow to detect a 50 reduction in mortality with a power of 90. The study s control group was therefore composed of 19,352 men which resulted in only 27 AAA-related deaths.
Results were presented as ratios with confidence intervals of 95. Men in the screening group were twice as likely to undergo elective surgery (p0.002). Screening of men aged 65 to 83 did not significantly decrease the mortality rate. Men aged 65 to 74 (normal target group) benefited more from screening then older men, the odds ratio between this subgroup and the control group being of 0.19 (p0.001). Results considered significant in the study were presented with adequate confidence intervals and p values, making them reliable at both ends of the interval.
The inclusion of older participants in this study, combined with the lack of exclusion criteria and low attendance rates confounded the results. This did however stress the importance of targeting the proper age group in order to benefit from screening. The study also suggested that invitations to screening may be better accepted if sent by the patient s practitioner rather than an unknown source.
Randomised clinical trial of screening for abdominal aortic aneurysm in women
The study s objective was to determine the benefit of screening women living in Chichester, aged 65 to 80 years of age, for AAA. This was a RCT comparing the outcome of patients in a screening group to a control population. No information was provided on the randomization process. Participants were apparently randomly allocated to either the screening group or the control group, which were of equal sizes and age-matched. The article describes no attempts to achieve blindness.
Only 3052 of the 4682 women invited to screening accepted. The article provides no explanation for the number of participants selected nor any description of the statistical analysis that the authors planned to use to evaluate the results. Results pertaining to death were obtained for all participants through the district Registrar of Births and Deaths. Only women with detected aneurysms were invited to follow-up exams. Four of the ten women in the screening group who died from ruptured AAA had had normal initial scans.
The results are presented as absolute values or percentages, with no reference to ratios, confidence intervals or p values. However, the results are very clear. Indeed, the prevalence of AAA-related mortalities was so low in both the screening and control groups that statistical evaluation was unnecessary to prove that screening did not reduce the rate of mortality.
The study discovered that women are affected by AAA later in life and that, contrary to men, a higher percentage suffer from AAA rupture after the age of 80 years. It would have been interesting for the study to include women over the age of 80, even though these patients are much less likely to be eligible for elective surgery.
Discussion
The study entitled Immediate Repair Compared with Surveillance of Small Abdominal Aortic Aneurysms compared survival rates in patients aged 59 to 70 years of aged, diagnosed with an AAA. Patients either underwent immediate surgical repair of their aneurysms or entered a surveillance program, with surgery reserved for certain situations. This study was well designed with proper attention put on randomization and blinding. However, the study was composed of a limited number of participants when compared to the other studies. Furthermore, participants were aged from 50 to 70 years of age, which is lower than the recommended screening age. According to the United Kingdom National Screening Committee (UKNSC), screening for AAA should be offered to men 65 years or older. By including younger individuals in their study, the authors may have biased the results. Indeed, younger patients would be healthier and more resistant to surgery. This study should have been designed according to the UKNSC screening criteria to better evaluate the proper target group and ensure the credibility and relevance of the results.
The Danish study entitled Screening for abdominal aortic aneurysms single centre randomised controlled trial was carried out on a larger scale. However, the lack of details provided concerning the randomization process is a major weakness as it may hide any number of biases. Furthermore, the participants were all chosen from the same county which may have been another cause for bias. Finally, the age limit of patients accepted into the trial was slightly lower than the age recommended by the UKNSC. This is another possible cause of bias that may have altered the results, making them less applicable to the United Kingdom.
The UK multicentre study was a very well designed study. The participants selected corresponded to the UKNSC criteria. Statistical analysis provided credible results and proved that screening the proper population of men for AAA reduced AAA-related deaths by 42. The study s main weakness was the use of death certificates to determine cause of death. Indeed, though all efforts were made to limit bias at this level, the retrospective approach may have altered some of the results. Finally, only a limited number of participants were questioned to determine the effects of screening on quality of life. This may also have biased the results.
The Western Australian study clearly demonstrates the importance of correctly choosing the screening population. In screening men younger than 65 years of age, no significant reduction in mortality rates was noted. However, when the data for the proper age group was evaluated alone, the results became significant. This study may have other sources of bias hidden in the group allocation process. Indeed, the article failed to describe the randomization process used.
Finally, the UK screening study performed on women was a weakly designed and poorly presented study. Whereas the UKNSC recommends screening men over the age of 65, this study screened women over the age of 65. As was predictable when considering the lower prevalence of AAA in women, the program showed that screening women was neither beneficial in saving live nor cost efficient. Furthermore, the results of this study are unreliable for many reasons. First, no description of the randomization process was provided. Second, the data was not provided with confidence intervals, p values or any mention of statistical analysis.
To summarize the results of these five studies, a proper screening program for AAA aimed at men over the age of 65 years will be beneficial in saving lives.
Statistical relevance
It is estimated that 0.5 to 3.2 of the population suffers from AAA (Hallet 2008) and that the disease is responsible for 2.1 of deaths in men over the age of 65 (MASS 2002). In the United States alone, this accounts for 9,000 deaths per year, 33,000 elective surgeries and 1400 to 2800 surgically related deaths (The Aneurysm Detection and Management Veterans Affairs Cooperative Study Group 2002).
Clinical aspects
Patients suffering from AAA rarely show clinical signs before the aneurysm ruptures (Lindholt 2005). When clinical signs are present, they tend to be discreet, depending on their size and growth rate. Though rare, rapidly growing aneurysms can be sensitive, especially before they rupture. In general, aneurysm patients that present clinical signs will complain of a constant pain in the lumbosacral region or will present an abnormal pulsating mass in the lower abdomen. In some cases, auscultation of the area will detect a systolic murmur (Hallet 2008). Unfortunately, the most commonly noted signs of AAA appear once the aneurysm has ruptured. Approximately half of these patients will die before arriving at a hospital and an additional 30 to 70 will die upon arrival (MASS 2002). This high mortality rate of 65 to 85 stresses the necessity for adequate screening of all patients at risk for AAA.
Diagnosis
AAA can be diagnosed through different imaging techniques. Ultrasonography is the most cost efficient, minimally invasive technique. Other possibilities include computed tomography or magnetic resonance angiography which can be useful in defining the aneurism s size and anatomy (Hallet 2008). In some cases, routine radiographs performed in search of other abnormalities may allow the practitioner to view a calcification of the affected arterial wall.
Treatment
A ruptured abdominal aortic aneurysm requires emergency surgery. Two surgical techniques exist. The first involves open laparotomy. This technique is associated with a 50 mortality rate. Endovascular stent grafting has a lower, though still considerable mortality rate of 20-30 (Hallet 2008). Mortality rates are high in both cases due to the precarious cardiovascular state these patients arrive in and to preexisting concomitant cardiovascular diseases. In cases where AAA is detected before rupture occurs, elective graft surgery may be proposed depending on the risk of rupture but also on the presence or not of contraindications for surgery. Indeed, because the elective surgery is associated with a 2-6 mortality rate, it is important to carefully select patients for this preventive surgery (MASS 2002). Conservative treatment consists of controlling risk factors such as hypertension.
Screening is an important measure in controlling AAA. Patients diagnosed with this ailment must be followed on a regular basis and informed of the possibility of elective surgery and of it s associated mortality rates. Ultrasonography is the technique of choice as it minimizes the patient s exposure to radiation. The following articles discuss different screening programs.
Immediate Repair Compared with Surveillance of Small Abdominal Aortic Aneurysms
The study clearly stated it s objective to compare the outcome of patients diagnosed with small AAA according to their treatment plan a) immediate surgical repair, or b) ultrasonographic or CT surveillance with surgery reserved for high risk aneurysms. Patients 50 to 79 years of age who had been diagnosed by CT with an AAA 4.0 to 5.4 cm in diameters within 12 weeks of the beginning of the study were included. A complete list of exclusion criteria was also provided in the article. The study was then correctly carried out as a RCT to avoid bias in treatment assignment, to facilitate blinding and to allow the use of the probability theory.
Patients were randomly assigned to one of two blocks (immediate repair or surveillance) by a computer generated random-number code. The study was composed of two equally sized groups of over 500 patients each. This allowed to statistically prove the results with a confidence level of 95. Randomization was stratified for each medical centre to ensure a proper balance of participants characteristics between groups.
The two groups were very similar at baseline. Considering the surgical nature of the treatment made it was impossible to conceal to which group patients had been allocated. To compensate for this, neither patients nor investigators had access to the outcome data during the study and the causes of all deaths were determined by a blinded outcomes committee. CT scans were also evaluated by a central laboratory. Though not perfectly blinded, all efforts were made to eliminate observer bias.
The status (dead or alive surgically repaired or not) of all 1136 patients was accounted for at the end of the study and the group related outcomes evaluated through intention-to-treat analysis. 92.6 of immediate-repair patients and 61.6 of the surveillance group patients had undergone surgery. It was unclear why the remaining 7.4 of immediate-repair patients had not undergone surgery. Data from surviving patients was collected through follow-up visits and final CT scans. Causes of death were determined by an outcomes committee.
Results were presented as a relative risk of death, either from AAA-related causes or all causes. No significant difference in outcome was noted between the immediate-repair group and the surveillance group.
A confidence level of 95 was reported for all results. The results are precise enough to encourage patients to follow a surveillance program rather than opting for immediate elective surgery for small aneurysms.
This study proved that immediate-repair was not associated with a higher survival rate but rather with more hospitalizations and therefore a higher cost than a surveillance program. Unfortunately, the majority of the patients included in this study were men. Though their is a higher prevalence of AAA among older men, the study would have been more interesting had it included more women, especially considering that women are at a higher risk for AAA rupture. The results of this study are therefore relevant for men but not necessarily for women.
Screening for abdominal aortic aneurysms single centre randomised controlled trial
The study was clearly focused on determining if AAA screening reduced mortality rates in Danish men (64 to 73 years old) from Viborg County. It was appropriately performed as a RCT, comparing a screening group to a control group. The means by which 12,639 men were randomly assigned to either the screening group (6333) or the control population (6306) are not described. Though the groups were balanced in numbers, little data is available making it impossible to judge if the two groups were statistically similar at baseline.
The patients were not blinded and were informed of the risk factors for AAA, possibly causing a behavior bias. All outcome data was analyzed by two blinded surgeons, thus eliminating observer bias. All 12,639 men who entered the trial were accounted for (dead or alive treated or not). No patients were transfered from their original group even though 1481 men from the screening group did not attend screening. Analysis of all outcome data was carried out on an intention-to-treat principal.
Patients in both groups were followed up for 0 to 69 months. Though the two groups were apparently similar for duration of follow-up, this still represents a large variation in data collection. Data on deaths and surgical procedures were collected by the same means in both groups.
The results are presented as relative risks with a confidence interval of 95 and appropriate p values. They are very precise with very low p values and acceptably small intervals. The results prove that screening significantly increases the number of men to undergo elective surgery and decreases the number of ruptured AAA as well as the number of AAA-related deaths. The benefits of screening on outcome are as clear at the lower confidence level as at the higher level.
All outcomes (relative risk of outcome, elective or emergency surgery and AAA rupture) were properly considered between the two groups. The study excluded women and gave very little information on it s population, thus making it difficult for readers to compare the study groups to any local populations. However, the discovery that only 352 men must be screened to save one life makes the results of this study particularly interesting, even considering the study s shortcomings.
The Multicentre Aneurysm Screening Study into the effect of abdominal aortic aneurysm screening on mortality in men a randomised controlled trial.
The study s clearly stated objectives were to evaluate the mortality rate (AAA-related and not) and the quality of life of men, aged 65 to 74 years living in the UK, screened for AAA in comparison with a control population. The study was therefore appropriately designed as a RCT with patients being randomly assigned to either a screening group or a control group. Patients were randomly allocated to one of the two groups through computer generated pseudorandom numbers. Randomization was then balanced at baseline for screening centre, age and social deprivation scores.
Attempts were made to carry out a blinded study. For instance, a data monitoring committee was responsible for evaluating death related information and the sponsors of the trial played no role in the study.
Results were not provided to the trial coordinators until all data had been collected. 80 of the men invited for screening were evaluated. Final outcome (dead or alive) was available for 99 of the men initially randomized.
Clinical follow-up was available for 81 of patients in whom screening had detected an aneurysm. Data was evaluated following a pre-specified plan. Death certificates were obtained through the Office of National Statistics and analyzed by a data monitoring committee. Surgical data was obtained through the medical centers and subsamples of each group completed identical tests to evaluate their quality of life following screening.
Before the trail began, it was determined that 66,000 men needed to be randomized to detect a 30 decrease in mortality with an 80 power of detection and a 5 significance level. This was based on approximately 115 ruptured AAA deaths in the control group. Over 70,000 men were therefore randomized which allowed to observe 91 deaths from ruptured AAA in the control population.
Results were presented as hazard ratios and corresponding absolute risks. AAA-related deaths were reduced in the screening group when compared to the control population (hazard ratio of 0.58). No significant change in mental health was noted in the men, regardless of their group and screening results. Outcome was evaluated with a 95 confidence level, a tight interval and a low p value of 0.0002. Mental health results were also subjected to low p values.
The financial aspect of screening was not discussed in this study. Though a reference was made to future publication of these results, a comment on the cost of screening would have been interesting.
Population based randomised controlled trial on impact of screening on mortality from aaa
The study aimed to discover if a screening program for AAA would reduce AAA-related mortality rates in Perth men aged 65 to 79 years. It was appropriately carried out as a RCT, comparing a study group to a control group. Participants were allocated to two equal sized groups balanced for age. No description of the randomization process is given making it impossible to evaluate if the process was appropriate or biased.
All attempts were made to eliminate observer bias. Australia s unified database system allowed for all data pertaining to the participants surgical procedures andor deaths to be accounted for at the end. Data was submitted to an intention-to-treat analysis. A blinded researcher evaluated death certificates for cause of death. When AAA was involved, a second blinded analysis was completed for confirmation.
All outcome data was collected through Australia s unified database system. Both groups suffered equal losses between randomization and screening. Low participation rates may have biased the results in contrast to other studies with stricter inclusion criteria and higher participation rates. It was estimated that a study involving a control group of 20,000 men experiencing about 55 AAA-related deaths would allow to detect a 50 reduction in mortality with a power of 90. The study s control group was therefore composed of 19,352 men which resulted in only 27 AAA-related deaths.
Results were presented as ratios with confidence intervals of 95. Men in the screening group were twice as likely to undergo elective surgery (p0.002). Screening of men aged 65 to 83 did not significantly decrease the mortality rate. Men aged 65 to 74 (normal target group) benefited more from screening then older men, the odds ratio between this subgroup and the control group being of 0.19 (p0.001). Results considered significant in the study were presented with adequate confidence intervals and p values, making them reliable at both ends of the interval.
The inclusion of older participants in this study, combined with the lack of exclusion criteria and low attendance rates confounded the results. This did however stress the importance of targeting the proper age group in order to benefit from screening. The study also suggested that invitations to screening may be better accepted if sent by the patient s practitioner rather than an unknown source.
Randomised clinical trial of screening for abdominal aortic aneurysm in women
The study s objective was to determine the benefit of screening women living in Chichester, aged 65 to 80 years of age, for AAA. This was a RCT comparing the outcome of patients in a screening group to a control population. No information was provided on the randomization process. Participants were apparently randomly allocated to either the screening group or the control group, which were of equal sizes and age-matched. The article describes no attempts to achieve blindness.
Only 3052 of the 4682 women invited to screening accepted. The article provides no explanation for the number of participants selected nor any description of the statistical analysis that the authors planned to use to evaluate the results. Results pertaining to death were obtained for all participants through the district Registrar of Births and Deaths. Only women with detected aneurysms were invited to follow-up exams. Four of the ten women in the screening group who died from ruptured AAA had had normal initial scans.
The results are presented as absolute values or percentages, with no reference to ratios, confidence intervals or p values. However, the results are very clear. Indeed, the prevalence of AAA-related mortalities was so low in both the screening and control groups that statistical evaluation was unnecessary to prove that screening did not reduce the rate of mortality.
The study discovered that women are affected by AAA later in life and that, contrary to men, a higher percentage suffer from AAA rupture after the age of 80 years. It would have been interesting for the study to include women over the age of 80, even though these patients are much less likely to be eligible for elective surgery.
Discussion
The study entitled Immediate Repair Compared with Surveillance of Small Abdominal Aortic Aneurysms compared survival rates in patients aged 59 to 70 years of aged, diagnosed with an AAA. Patients either underwent immediate surgical repair of their aneurysms or entered a surveillance program, with surgery reserved for certain situations. This study was well designed with proper attention put on randomization and blinding. However, the study was composed of a limited number of participants when compared to the other studies. Furthermore, participants were aged from 50 to 70 years of age, which is lower than the recommended screening age. According to the United Kingdom National Screening Committee (UKNSC), screening for AAA should be offered to men 65 years or older. By including younger individuals in their study, the authors may have biased the results. Indeed, younger patients would be healthier and more resistant to surgery. This study should have been designed according to the UKNSC screening criteria to better evaluate the proper target group and ensure the credibility and relevance of the results.
The Danish study entitled Screening for abdominal aortic aneurysms single centre randomised controlled trial was carried out on a larger scale. However, the lack of details provided concerning the randomization process is a major weakness as it may hide any number of biases. Furthermore, the participants were all chosen from the same county which may have been another cause for bias. Finally, the age limit of patients accepted into the trial was slightly lower than the age recommended by the UKNSC. This is another possible cause of bias that may have altered the results, making them less applicable to the United Kingdom.
The UK multicentre study was a very well designed study. The participants selected corresponded to the UKNSC criteria. Statistical analysis provided credible results and proved that screening the proper population of men for AAA reduced AAA-related deaths by 42. The study s main weakness was the use of death certificates to determine cause of death. Indeed, though all efforts were made to limit bias at this level, the retrospective approach may have altered some of the results. Finally, only a limited number of participants were questioned to determine the effects of screening on quality of life. This may also have biased the results.
The Western Australian study clearly demonstrates the importance of correctly choosing the screening population. In screening men younger than 65 years of age, no significant reduction in mortality rates was noted. However, when the data for the proper age group was evaluated alone, the results became significant. This study may have other sources of bias hidden in the group allocation process. Indeed, the article failed to describe the randomization process used.
Finally, the UK screening study performed on women was a weakly designed and poorly presented study. Whereas the UKNSC recommends screening men over the age of 65, this study screened women over the age of 65. As was predictable when considering the lower prevalence of AAA in women, the program showed that screening women was neither beneficial in saving live nor cost efficient. Furthermore, the results of this study are unreliable for many reasons. First, no description of the randomization process was provided. Second, the data was not provided with confidence intervals, p values or any mention of statistical analysis.
To summarize the results of these five studies, a proper screening program for AAA aimed at men over the age of 65 years will be beneficial in saving lives.
Categories:
0 comments:
Post a Comment