Summary

Objective: To assess the effect of coexisting medical morbidity on the epidemiology of tuberculosis in Hong Kong.

Design: A cross-sectional observational study.

Subjects: All tuberculosis patients notified to the Department of Health in the month of August 1999.

Main outcome measures: For all these patients, a questionnaire was sent to the doctor who notified the case. It was a standardised data collection form asking for any coexisting medical conditions, which are considered as predisposing factors for tuberculosis.

Results: Out of 622 cases of tuberculosis notified in August 1999, there was no reply for 28 cases (4.5%). Among the remaining 594 cases, co-morbidities which could predispose to the development of tuberculosis were found in 155 patients (26%). Diabetes mellitus was the most prevalent condition, with an overall prevalence of 12.1%. Malignancies ranked second in the list, with a prevalence of 4.8%. Only one case of HIV infection was reported in the tuberculosis registry in that month.

Conclusion: These data confirm and extend the results of other studies, showing that substantial proportions of tuberculosis patients in Hong Kong are elderly and have significant comorbidities. The prevalence of HIV infection associated with tuberculosis is still low in Hong Kong. The high percentages of coexisting medical morbidities in tuberculosis patients add an extra burden to the provision of care to these patients in the health care system.

Keywords: tuberculosis, epidemiology, comorbidity

摘要

目的：評估結核病人中共存的其他疾病的情況及其對 香港結核病的流行病學程度的影響。

設計：這是一個截面式觀察性研究。

對象：所有於一九九九年八月份所有向衛生署呈報的結核病病人。

測量內容：衛生署胸肺科會向每個個案的主診醫生寄出的問卷，查詢關於病人其他疾病的資料，包括一些 被認為可能會誘發結核病的疾病。

結果：一九九九年八月，共有 622宗結核病個案呈報給衛生署。在回收到的594份問卷中，有 155個病人（26%）同時患有可能誘發結核其他疾病。其中以糖 尿病佔首位，比率為12.1%；癌症排行第二，比率為4.8%；而愛滋病在該月份的結核病呈報個案中則只有 一宗。

結論：這些結果進一步證實及支持其他研究的結論， 顯示香港結核病人有相當部份為老年人以及患有其他疾病。愛滋病患者在香港結核病者中仍是屬於少數。 結核病病人共患有其他疾病的比例較高，因此醫療系統所承受的負擔亦相應增加。

主要詞彙：結核病，流行病學，共同罹病率

---

Introduction

Tuberculosis (TB) is a leading cause of morbidity and mortality worldwide.¹ In Hong Kong, the incidence of TB was very high in the late 1940s, at close to 700/100,000. With the effective control programme of the Government Tuberculosis and Chest Service of Hong Kong, and the progressive improvement in the socio-economic situation, the notification rate of TB decreased to 100.9/100,000 in 1995. Since then, there has been an increase in the incidence of TB to 101.0 in 1996, 109.0 in 1997, 117.3 in 1998, 113.7 in 1999 and 113.7 in 2000.² A similar trend of increase in rate has been observed in a number of other countries.^3,4 While the rise in incidence of TB may be artificial as a result of the increased awareness of the medical profession of the importance of notification,^2,5 the increase could also be real. Even if the rise in the notification rate were artificial, the lack of a decline in the notification rate of TB over the past 10 years has to be explained. Various reasons have been proposed, including:

1. a change in the disease agent such as emergence of multi-drug resistant TB;
2. a change in the disease vector, such as an increase in overcrowding environment in Hong Kong and widespread use of air-conditioning in the indoor environment;
3. a change in the host such as increasing number of HIV infected patients, an ageing population with prolonged survival of people with medical illnesses as that can predispose to the development of TB disease.

The purpose of this study is to assess the effect of the prevalence of various medical conditions on the epidemiology of TB in Hong Kong.

Study design, data collection and analysis Patients TB is a statutorily notifiable disease in Hong Kong. All TB patients (pulmonary and extrapulmonary) notified to the Department of Health in the month of August 1999 were included.
Study design A cross-sectional observational study was employed. Definitions used were based on that from the International Union Against Tuberculosis and Lung Diseases:6
	A case of TB was defined as one who had tubercle bacilli visible on microscopic examination of sputum (smear-positive) or on culture examination if smear was negative; or in the absence of positive bacteriology, those with clinical and radiological features compatible with TB. Pulmonary cases
	-	those with TB of the lungs including those who were sputum smear-positive and those who were sputum smear-negative (provided a minimum of three sputum examinations had been performed).
	Extrapulmonary cases
	-	all other patients, including those with TB pleurisy and miliary TB.
A questionnaire was sent to the doctor who notified the case, asking for comorbidities that could predispose to TB. Risk factors for development of TB include co-infection with HIV, abuse of alcohol and other illicit drugs, diabetes requiring oral or parenteral hypoglycaemic therapy, end stage renal failure, haematological diseases such as leukaemia, lymphoma, or other disorders of the reticuloendothelial system, silicosis, and drug-induced immunosuppression associated with corticosteroids, chemotherapy, or other immunomodulating drugs used to treat patients with organ transplant and collagen vascular diseases. Analysis The questionnaire was then collected and analysed by the Statistics Unit of the Department of Health. The data were entered into Epi-Info database, version 67 with accuracy checked and then analysed.

Results

There were 622 cases of TB notified in August 1999. Standard questionnaires were sent to doctors who notified the case. There was no reply for 28 cases, with a dropout rate of 4.5%. Male to female ratio was 1.7:1. Age group distribution is shown in Table 1. It can be seen from the table that the age distribution of the study population is similar to the TB population of 1999. The predominant age of presentation was in the 7th and 8th decades, meaning that TB in Hong Kong is most prevalent in the older age groups. Co-morbidities were found in 26% (i.e. 155) of all patients. As shown in Chart 1, patients with coexisting medical comorbidities were concentrated in the elderly population. For patients under 60, 17% had co-existing medical comorbidities. This contrasts with patients above 60, for whom 41% had co-existing medical comorbidities.

Table 1: Age distribution of the study population
Age group	Frequency	% in the present study population	Overall % in the TB notification registry of 1999
0 to 9	5	0.8	0.6
10 to 19	25	4.0	3.8
20 to 29	76	12.2	12.7
30 to 39	68	10.9	13.3
40 to 49	88	14.1	13.8
50 to 59	75	12.1	11.7
60 to 69	107	17.2	16.3
70 to 79	108	17.4	17.3
Over 80	70	11.3	10.4

Chart 1: Comorbidities among TB patients

The prevalence of various medical conditions is tabulated in Table 2. Among these, diabetes mellitus was the most prevalent condition in both groups, accounting for 5.6% of patients under 60 and 18.6% of patients above 60. Malignancies (including leukaemia and lymphoma) ranked second in the list, occurring in 2.4% of patients under 60 and 8% of patients over 60. The prevalence of co-infection with HIV was low in the study group with only one HIV seropositive patient.

Table 2 also shows the differences in medical comorbidities in patients notified from community-based chest clinics versus those from hospitals. In general, the prevalence of comorbidities among patients reported from hospitals was double that from chest clinics. Furthermore, patients notified from hospitals tended to have more than one medical comorbidity (5.7% of patients notified by hospitals had more than one medical comorbidity, versus 0.3% of those notified by chest clinics).

When patients were classified into pulmonary and extrapulmonary groups (Table 2), the prevalence of medicalcomorbidities was similar in the two groups except a higher prevalence of chronic renal failure in extrapulmonary TB and a higher prevalence of malignancies, pneumoconiosis, drug addiction and alcoholism in pulmonary TB.

Table 2: Comorbidities among tuberculosis (TB) patients (classified according to age, source of notification, pulmonary and extra pulmonary TB)
Comorbidity	% among patients aged <60	% among patients aged >60	% among patients from chest clinics	% among patients from hospitals	% among patients with Pulmonary TB	% among patients with Extrapulmonary TB	Overall %
	total=337 no reply=8	total=285 no reply=20	total=298 no reply=0	total=324 no reply=25	total=556 no reply=25	total=66 no reply=3
Diabetes mellitus	5.6	18.6	8.7	15.5	12.4	9.2	12.1
On steroid	0.9	1.8	1	1.7	1.2	1.6	1.3
Chronic renal failure	1.2	4.9	0.3	5.7	2.1	7.9	3.0
On cytotoxic drug	0.0	0.4	0	0.3	0.2	0	0.2
Leukemia/lymphoma	0.3	1.1	1	0.3	0.75	0	0.7
Other malignancy	2.1	7.0	2	6.4	4.1	0.04	4.2
Alcoholism	1.78	1.4	0	2.7	1.9	0	1.7
Drug addiction	1.48	0.4	0	2	1.1	0	1.0
Pneumoconiosis	1.2	1.8	1	6	1.7	0	1.52
Other medical conditions	3.0	3.5	1	6.4	3.8	3	3.70
Any medical condition	17.7	40.3	15.9	40.8	28	26	26
Other medical conditions include anorexia nervosa, HIV infection, post heart transplant, SLE, treatment with cyclosporin A

Discussion

The prevalence of various coexisting medical conditions in TB patients varies in different series. There are numerous studies in the literature showing the different annual incidence of development of TB in a number of medical conditions such as diabetes,^8,9 renal failure,¹⁰ malignancy¹¹ etc. However, studies showing the prevalence of various diseases among TB patients in a locality are scarce. This study is the first in Hong Kong to show the prevalence of various disease comorbidities in TB patients.

Among various medical comorbidities, diabetes mellitus was most common, accounting for 5.6% of patients below 60 and 18.6% patients above 60, with an overall prevalence rate of 12.2% and a standardised age adjusted prevalence rate of 5.7% (compared with an overall general prevalence of diabetes mellitus in Hong Kong of around 4.6% and an age-adjusted rate of 7%).^12,13 Our figure is quite comparable to that of Japan,^14,15 which showed a prevalence of 13.3 to 21.3%. (With general prevalence of 9.6-11.9% in people aged 40 years or over).¹⁶ On the other hand, Poland¹⁷ and Nigeria¹⁸ had a much lower prevalence rate of 2% and 5% respectively (with general prevalence of 3.2%¹⁹ in Poland). Malignancies ranked second among the comorbidities in our TB patients, accounting for 2.4% of patients below 60, 8.1% of patients above 60 and an overall prevalence of 4.2%. Studies in Saudi Arabia²⁰ showed that 10% of miliary TB patients had coexisting malignancy, while in Poland,²¹ 17% of TB patients had a history of lung cancer.

Drug addiction and alcoholism occurred in only 1.6% and 1%, respectively, of our patients. In other studies, alcoholism was a significant factor, accounting for about 10% of TB cases in Poland²² and Russia,²³ 23% in Hungary,²⁴ and 47% in Central Los Angeles.²⁵ Drug users were also common in studies in the United States, with a prevalence of 20% in New York City²⁶ and 36% in Central Los Angeles.¹²

Concomitant infection with HIV was uncommon in our patient population. In our study, only one patient was found to be HIV positive in the month of August 1999. This contrasts sharply with the United States where the HIV epidemic has played a major role in the resurgence of TB.²⁷

In an audit done of TB service in Hong Kong in 1994,²⁸ a set of 454 consecutive notifications of patients dated between December 1st to 31st of 1994 was studied. The sample size of that study is smaller than this study, reflecting an increased notification rate of TB in the more recent year. In the earlier report, 2012 episodes of care were identified for the 454 patients. 118 episodes (2%) were principally for comorbidities other than TB. In 1220 episodes (55.7%) of care, the patients reported that the comorbidity was an active ongoing problem. In 579 episodes of care (48.3%) the comorbidities directly affected the management of TB in these patients. This reflects the severity of the burden of care of the comorbidities in TB patients. In fact, in their audit, Hedley et al stated that comorbidities, complications of TB and treatment side-effects were associated with much higher levels of need and were the most important cost driver in the TB service. In their audit, diabetes mellitus again was the most frequently reported comorbidity (15.3% of all episodes reported).

There are several major limitations for our study. With its cross-sectional observational design, our data may reflect only a limited glimpse of the medical comorbidities in TB patients. This study looks only at the prevalence of medical comorbidities. However, the effects of these comorbidities on the patients' outcomes have yet to be analysed. One other study²⁹ has shown that presence of medical comorbidities such as end stage renal failure, respiratory failure, age >60 and immunosuppression increases the mortality of TB patients (with an adjusted odds ratio varying from 3.5 to 7.0). Further research in this area, with a new set of data collection forms devised by the Chest Service, has been implemented, with the aim of shedding more light on this area in the future.

Conclusion

With the above caveats in mind, the data presented herein should provide us with a picture of comorbidities of TB patients in our locality. HIV infection among TB patients is still low in Hong Kong. TB among the elderly has become an increasingly important problem in recent years. All doctors should watch out for coexisting medical morbidities in this group of patients.

Key messages

1. Tuberculosis is still an important communicable disease with high prevalence in Hong Kong.
2. Its incidence has increased among the elderly population.
3. A significant proportion (26%) of tuberculosis patients suffer from coexisting medical comorbidities that can predispose to the development of tuberculosis. Among these, diabetes mellitus is the most common (12%).
4. General physician should be alerted to watch out for coexisting medical comorbidities in tuberculosis patients, especially the elderly.

---

---

References