Estimates of Cancer Mortality in Hanoi and Ho Chi Minh City, Viet Nam in the 1990s

As cancer mortality data is not available, a study regarding the real problem of cancer mortality is timely and urgent in Viet Nam. Therefore the aim of the present study was to calculate cancer mortality in the city of Hanoi and Ho Chi Minh. The correlation between cancer mortality to incidence ratios and relative survival probabilities for 23 cancer sites was estimated according to SEER (1973-97), then cancer mortality was calculated from the cancer incidence and cancer survival for 25 cancer sites in each city. Cancer mortality rate for all cancer sites except skin (ASR per 100,000) was 103.9 for males and 52.4 for females in Hanoi, and 93.7 for males and 60.7 for females in Ho Chi Minh. For males, the five most common cancer deaths were cancers of the lung, liver, stomach, colon/rectum, and nasopharynx in both Hanoi and Ho Chi Minh. For females, cancer death in the cervix was uncommon in Hanoi but the most common site in Ho Chi Minh (ASR 2.2 VS. 14.2 per 100,000). The present findings are the first results of cancer mortality from Viet Nam and should be useful for further cancer control programs there.


Le Tran Ngoan, Tetsuya Mizoue and Takesumi Yoshimura
As cancer mortality data is not available, a study regarding the real problem of cancer mortality is timely and urgent in Viet Nam.Therefore the aim of the present study was to calculate cancer mortality in the city of Hanoi and Ho Chi Minh.The correlation between cancer mortality to incidence ratios and relative survival probabilities for 23 cancer sites was estimated according to SEER (1973-97), then cancer mortality was calculated from the cancer incidence and cancer survival for 25 cancer sites in each city.Cancer mortality rate for all cancer sites except skin (ASR per 100,000) was 103.9 for males and 52.4 for females in Hanoi, and 93.7 for males and 60.7 for females in Ho Chi Minh.For males, the five most common cancer deaths were cancers of the lung, liver, stomach, colon/rectum, and nasopharynx in both Hanoi and Ho Chi Minh.For females, cancer death in the cervix was uncommon in Hanoi but the most common site in Ho Chi Minh (ASR 2.2 VS. 14.2 per 100,000).The present findings are the first results of cancer mortality from Viet Nam and should be useful for further cancer control programs there.J Epidemio/, 2002 ; 12 : 179-187 Viet Nam, cancer mortality, survival, population-based-cancer-registry BACKGROUND Mortality statistics are commonly obtained from death certificate, however, death certificate are not available in Viet Nam at present 1,2).From 1995-98, the average annual number of deaths that occurred in hospitals from cancer was only 489 throughout the country (hospital based cancer registry).From these registered numbers, the proportion of cancer deaths was 0.1% of all causes of deaths and 1.5% among new cases of cancer [3][4][5][6].This proportion of deaths from cancer in Viet Nam was much lower than that of the estimated data for developing countries (0.1% VS. 9.0%) in the 1990s 7).The reason for this uncommon observation data for cancer mortality is that, only a small number of deaths occurred in hospitals nationwide (Annual cancer death number 489) and this caused serious biased data of cancer mortality in Viet Nam 8,7).This underestimation of cancer mortality dose not reflect the real public health problems of cancer morbidity and mortality nationwide.
Therefore, a study regarding the real problem of cancer mortality is timely and urgent in Viet Nam.The aim of the present study was to calculate cancer mortality in the two cities of Hanoi and Ho Chi Minh in the 1990s.

MATERIALS AND METHODS
The geographical unit in the present study was two cities: Hanoi and Ho Chi Minh.Cancer mortality was calculated based on 7 age groups from 0-14 to 65+ for given cancer sites.
The number of cancer sites examined in the present study was  1).In the city of Ho Chi Minh, the cancer survival data was available for cervical cancer only, therefore, the survival rate for the other 11 cancer sites was referenced from the results in Hanoi and applied to the survival rate in Ho Chi Minh.The 5-year survival rate for the remaining 13 cancer sites and other cancer sites except for skin was referenced from previous results for developing countries and applied to the 5-year survival rate of the 13 cancer sites in both Hanoi and Ho Chi Minh 11) Regarding the relative cancer survival between males and females, the estimated 5-year relative cancer survival for developing countries was seen to be slightly higher in males for cancers of the mouth and pharynx, stomach, larynx, and leukemia in comparison with those in females.That was about equal for cancers of the colon/rectum, melanoma of skin, bladder, kidney, and lymphoma but slightly lower for cancers of the esophagus, liver, pancreas, and lung when compared to those in females 12).Therefore, we decided to use the  4): 113 that was already excluded from the present analysis.For the other remaining cancer sites, the number of follow-up subjects lost was not stated.

Statistical Methods:
The strength of the relationship between cancer mortality to incidence ratios and relative survival probabilities after each given time period, 5, 3, and 1 year since the date of diagnosis was measured.Pearson Correlation Coefficients were calculated by using SAS software on a personal computer 15).The relationship between variables (M / I ratios and relative survivals: Si) was obtained as (M / I ratios) = a + b * (Si), then we have M = I * [(a + b * (Si)], where "a" and "b" was the Parameter Estimate of the intercept and the survival, respectively.

The Correlation between the Cancer Incidence and Mortality:
The present method of estimates of cancer mortality from incidence has been introduced in previous studies 11, 12).We have applied this method by using the latest data of SEER from 1973-97 13).Since the survival rate was available in Hanoi and Ho Chi Minh at 5, 3, and 1 year, the scatter diagram of data and a trend line for the relationship between mortality per incidence ratios and 5, 3, 1 year relative survival probabilities was calculated, where y = M : I ratio, x = S(5), S(3), and S(1), respectively.The equations are Y = -0.8717* X + 0.9219 (R2 = 0.9533), Y = -0.8709* X + 0.9648 (R2 = 0.9574), and Y = -0 .9554* X + 1.1468 (R2 = 0.9229) for 5, 3, and 1 year relative survival probabilities, respectively.Figure 1 shows the scatter diagram of data and a trend line for the relationship between mortality per incidence ratios and 1-year relative survival probabilities.

RESULTS
In the male population of Hanoi City during the 3-year period from 1991-93, the number of cancer deaths was 2,506, and the annual mortality rates were 80.3 and 103.9 per 100,000 (crude and ASR rate, respectively).More than one-fourth of cancer deaths was due to lung cancer (26.7%).The second most common cancer death was liver cancer (16.9%), followed by stomach cancer (14.7%) and nasopharygeal cancer (7.1%) (Table 2).For females, the number of cancer deaths was 1,513, and the annual mortality rates were 46.9 and 52.4 per 100,000 (crude and ASR, respectively), the most common cancer death was stomach cancer (15.1%), followed by breast cancer (14.1%), lung cancer (9.8%), and liver cancer (7.9%) (Table 2).
In the male population of Ho Chi Minh City during the 2year period from 1995-96, the number of cancer deaths was 2,960, and the annual mortality rates were 64.3 and 93.7 per 100,000 (crude and ASR rate, respectively).The most common cancer death was liver cancer (23.7%), followed by lung cancer (20.9%), stomach cancer (12.8%), and colon/rectum cancer (7.2%).For females, the number of cancer deaths was 2,632, and the annual mortality rates were 52.3 and 60.7 per 100,000 (crude and ASR rate, respectively).Nearly one-fourth of cancer deaths was due to cervical cancer (22.2%), followed by lung cancer (9.3%), stomach cancer (9.2%), colon/rectum cancer (9.1%), breast cancer (8.4%), and liver cancer (8.4%).
1-year relative survival probabilities Figure 1.Scatter diagram of data and a trend line for the relationship between mortality/incidence ratios and 1-year relative survival probabilities of SEER for 23 cancer sites.All cancer sites except for skin: Cancer mortality was higher among males (ASR 103.9 VS. 93.7 per 100,000) but lower among females in Hanoi (ASR 52.4 VS. 60.7 per 100,000) when compared to those in Ho Chi Minh.A fairly higher annual mortality rate was observed for the male age groups 35-44, 45-54, and 55-64 but a much lower annual mortality rate was observed for the female age groups 55-64 and 65+ in Hanoi when compared to those in Ho Chi Minh (Figure 2).However, for both sexes, cancer mortality for all sites except for skin was seen to be similar, that is, about 75 per 100,000 (ASR) and 60 per 100,000 (crude) in both Hanoi and Ho Chi Minh.
Lung cancer: For males, lung cancer was the most common cause of death from cancer in Hanoi, the second most common cause in Ho Chi Minh (ASR 29.3 and 20.8 per 100,000, respectively).For females, lung cancer was the third most common cause of death from cancer in Hanoi and the second most common cause in Ho Chi Minh (ASR 5.3 and 5.8 per 100,000, respectively (Table 2, Table 3).Sex ratio (M: F) was 5.5 in Hanoi and 3.6 in Ho Chi Minh.For both sexes, cancer death from lung cancer was the most common cause of death from cancer in Hanoi and the second most common cause in Ho Chi Minh (20.3% and 16.5% of all cancer sites in each city, respectively. Liver cancer: For males, liver cancer was ranked as the second most frequent cancer in Hanoi but the most frequent cancer in Ho Chi Minh (ASR 17.3 and 21.8 per 100,000, respectively).For females, it was the fourth most frequent cancer in Hanoi and the sixth most frequent cancer in Ho Chi Minh (ASR 4.2 and 5.1 per 100,000, respectively) (Table 2, Table 3).Sex ratio (M : F) was found to be about 4.0 both in Hanoi and Ho Chi Minh.For both sexes, death from liver cancer was the third most common cancer in Hanoi and the most common cancer in Ho Chi Minh (13.5% and 16.5% of all deaths in each city, respectively).
Stomach cancer: Stomach cancer mortality rate was the third most frequent cancer in both Hanoi and Ho Chi Minh in males, (ASR 15.8 and 12.5 per 100,000, respectively).For females, stomach cancer death was the most frequent cancer death in Hanoi and the third most frequent cancer in Ho Chi Minh (ASR 7.9 and 5.7 per 100,000, respectively) (Table 2, Table 3).Sex ratio (M : F) was about 2.0 in both Hanoi and Ho Chi Minh.For both sexes, death from stomach cancer was the second most common cause of death from cancer in Hanoi and the third most common cause in Ho Chi Minh (14.9% and 11.1% of all deaths in each city, respectively).
Nasopharygeal cancer: For males, nasopharygeal cancer was the fourth most common cause of death from cancer in Hanoi and the fifth most common cause of death from cancer in Ho Chi Minh (ASR 7.1 and 3.6 per 100,000, respectively).For females, it was the sixth most common cause of death from cancer in Hanoi and an uncommon cause of death from cancer in Ho Chi Minh, (ASR 3.0 and 1.0 per 100,000, respectively) (Table 2, Table 3).Sex ratio (M : F) was lower in Hanoi (2.4) than in Ho Chi Minh (3.6).
Colon/rectum cancer: In Hanoi, Colon/rectum cancer was the fifth most common cause of death from cancer in both Age groups Figure 2. Annual cancer mortality rate by age group in Hanoi and HCM for all cancer sites except for skin in males and females sexes (ASR 6.8 and 3.8 per 100,000, respectively).In Ho Chi Minh, it was the fourth most common cause of death from cancer in both males and females (ASR 6.7 and 5.6 per 100,000, respectively) (Table 2, Table 3).
Breast and cervical cancer: Breast cancer was the second and fifth most frequent cause of death from cancer in Hanoi and Ho Chi Minh, ASR 7.6 and 5.1 per 100,000, respectively.
Cancer death from cervical cancer was an uncommon cause of death from cancer in Hanoi but it was the most frequent cause of death from cancer in females in Ho Chi Minh (ASR 2.2 and 14.2 per 100,000, respectively) (Table 2, Table 3).A much lower annual mortality rate in Hanoi for all specific age groups than in Ho Chi Minh was seen and the highest annual mortality rate in Ho Chi Minh was seen for the age group 55-64 for cervical cancer (Figure 3).However, the present study may have some potential limita-tions.The relative survival rate was always larger than the observed rate for the same group of patients 13).Therefore, the number of cancer deaths estimated from observed survival may be, at least in part, overestimated for 12 cancer sites in the present study, which were calculated based on the data of hospital based cancer survival from Viet Nam.These data of observed survival were calculated on the basis of inpatient medical records.The patients with a serious advanced stage of incurable cancer that was diagnosed at outpatient clinics may not be included.Another limitation is data of hospital based cancer survival for 12 cancer sites from Viet Nam and the estimated data for another 13 cancer sites referenced from developing countries.Accuracy of hospital based cancer survival data in Viet Nam is very limited due to the small number of study subjects (Cancers of ovary, testis) and large number of follow-up subjects lost (Cancers of cervical uteri in Ho Chi Minh and penis) (Table 1).In addition estimated data from developing countries that is applied in the present study may not reflect the real problem in Viet Nam.We believe that our population based cancer survival study will provide a better database to calculate cancer mortality in the near future.
In spite of the limitations of the data sources, the present findings have indicated cancer mortality data in the city of Hanoi and Ho Chi Minh that should be useful for further cancer control programs in Viet Nam in particular and in Asia regions in general.

DISCUSSION
The present findings are the first results of cancer mortality from Viet Nam for the population of Hanoi and Ho Chi Minh in the 1990s.The annual crude mortality from all causes in Hanoi was about 620 per 100,000 in 1991-93 for both sexes 10) Also for both sexes, the present results have shown that the crude annual mortality rate from cancers at the same time in Hanoi was 63.3 per 100,000.Therefore, the proportion of deaths from cancers was 10.2% of all causes of deaths.The present study findings also present the number of deaths from cancer and its mortality rate for 25 cancer sites in both Hanoi and Ho Chi Minh in males and females.These findings also help us to compare cancer mortality between Hanoi and Ho Chi Minh.Since the 1950s in both Hanoi and Ho Chi Minh City, cancer treatment and a follow-up for cancer survival in general and cervical cancers in particular has been established 16, 17) Following the experiences of these researchers (Hien and Hoanh et al), cancer treatment and a follow-up has been done well for the 12 cancer sites used in the present study.

Table 2 .
Number of cancer deaths by age groups in males and females in Hanoi from 1991-93.
*World population

Table 3 .
Number of cancer deaths by age group in males and females in HCM from 1995-96 .
*World population Age groups Figure3.Annual cancer mortality rate by age group in Hanoi and HCM for breast and cervical cancers .