TuberculosisWith the advent of effective chemotherapy and surveillance techniques, the total number of cases and resultant mortality from tuberculosis decreased steadily until the early 1970s, but has not changed appreciably since that time; however, tuberculosis still causes approximately 2,000 deaths per year in the United States, making it the leading cause of death among 38 communicable diseases for which data on mortality are reported to the Center for Disease Control. In 1980, deaths from tuberculosis exceeded the combined total for the other 37 communicable diseases. The mortality in patients with tuberculosis continues to cause concern, but there has been little systematic evaluation of the actual cause of death in these patients. Hence, we have assessed the cause of death in 41 consecutive patients dying with active tuberculosis at the San Antonio State Chest Hospital.

Materials and Methods

The medical records and postmortem reports of all patients who died with active tuberculosis at the San Antonio State Chest Hospital between January 1979 and December 1983 were reviewed. No patients were excluded. From each record, information was extracted concerning critical demographic parameters including age, sex, Of the 41 patients, 34 were male and seven were female patients. The site of tuberculosis was pulmonary in 39 patients and extrapulmo-nary in three patients.

Active tuberculosis was defined as a positive culture within four weeks of death. The extent of tuberculous disease was assessed through the evaluation of postmortem and detailed roent-genographic reports, noting unilateral or bilateral pulmonary involvement, cavitary or noncavitary disease, and the number of lobes involved and the presence of extrapulmonary disease. Appropriate therapy was defined as the prompt institution of at least two drugs to which the organism was susceptible, without difficulty with compliance or undue delays in therapy due to adverse reactions to the drugs.

The cause of death was determined by specific criteria. Death was directly attributed to tuberculosis when there was involvement of the central nervous system with positive cultures at autopsy, overwhelming disease (progressive pulmonary tuberculosis, without respiratory failure, but with evidence of negative nitrogen balance), disease with a totally drug-resistant organism with no other apparent cause of death, massive pulmonary hemorrhage immediately before death or respiratory failure (oxygen pressure less than 50 mm Hg; carbon dioxide tension greater than 50 mm Hg with progressive respiratory acidosis) with no other apparent cause of respiratory failure. Nontuberculous causes of death were assigned based onclinical presentation, clinical course, and postmortem findings. Special note was also made when medical therapy aggravated respiratory function or clotting abnormalities.


Over the five-year period, there were 2,937 tuberculosis-related admissions at the San Antonio State Chest Hospital. Forty-one patients died with active tuberculosis. The mean age of those dying with central nervous systemactive tuberculosis was 59 years (range, 1 to 92 years). The majority of patients (36) were aged 40 years or older. Male patients outnumbered female patients (34/7). All but three of the patients had only pulmonary disease (one renal and two meningitis).

Twenty (49 percent) of the patients were considered to have died as a direct result of tuberculosis. Seven died of overwhelming disease, four of massive hemoptysis, six of respiratory failure, two of involvement of vital organs, and one from progressive disease with a totally drug-resistant organism (Table 1).

The mean age of the seven patients dying as a result of overwhelming disease was 70 years (range, 56 to 89 years). The duration of therapy in these patients prior to death varied from one week to two years, with five out of seven having received therapy for less than five weeks. Therapy was considered appropriate in all of these patients. Two were considered to be potentially immunosuppressed, and hypoalbuminemia was found in all four patients in whom determinations had been performed. Roentgenographic findings of bilateral and at least three-lobe involvement were seen in six patients, and cavitary disease was present in four.

The four patients dying from massive hemoptysis had a mean age of 48 years (range, 35 to 64 years). Only one of these patients received appropriate therapy. One had a totally drug-resistant organism, having received multidrug therapy, and the other two patients were noncompliant. All patients dying of hemoptysis had bilateral cavitary pulmonary disease with two or more lobes involved. None of the patients had an underlying medical condition, nor were they receiving a medication which could potentially alter hemostasis.

The six patients dying of respiratory failure had a mean age of 53 years (range, 37 to 79 years). All were men. Therapy was considered appropriate in each case. The length of therapy prior to death ranged from two days to one year, with the exception of one patient who was receiving second-line drug therapy due to first-line drug resistance and had been receiving therapy for more than two years. No patient was receiving therapy which would potentially suppress respiration or had any respiratory failuresignificant underlying pulmonary disease. Albumin levels were 2.5 mg/dl or less in the four patients in whom they were determined. Four patients had bilateral cavitary disease with three or more lobes involved, one patient had bilateral noncavitary disease, and another had unilateral cavitary disease.

Two patients died of involvement of vital organs due to tuberculous meningitis. One of these patients also had miliary pulmonary disease and involvement of lymph nodes, spleen, liver, bone marrow, kidneys, and adrenals. Both were male patients, aged 7 and 62 years. The diagnosis in both cases was made after death. The diagnosis of tuberculosis was suspected before death in one patient, and he was empirically placed on therapy with isoniazid, rifampin, and streptomycin and had received therapy for one week prior to death. Results of culture not available before death revealed the organism to be resistant to isoniazid and streptomycin.

One patient died of tuberculous disease with a totally drug-resistant organism. He had received multiple drugs over a two-year period and had progressive pulmonary involvement (four lobes at the time of death). The exact cause of death was undetermined.

Twenty-one patients (51 percent) died of nontuber-culous causes in the setting of active tuberculosis. Pulmonary embolus was the cause of death in five patients. Two patients each died from documented dysrhythmias and from acute myocardial infarction. Massive hemoptysis accounted for one death which occurred in a patient with a pulmonary carcinoma. Respiratory failure due to chronic obstructive pulmonary disease (COPD) and bacterial pneumonia each accounted for one death, gastrointestinal bleeding for three deaths, bacterial sepsis for two deaths, and cerebrovascular accident for two deaths. The cause of death could not be determined in two patients. One of these died suddenly of a “cardiac arrest.” He had cavitary disease and had been on appropriate therapy for two months. The other patient also had cavitary pulmonary disease with progressive respiratory deterioration. Although samples of sputum were persistently negative,Mycobacterium tuberculosis was grown from the urine. The serum level of albumin was abnormal in all nine patients in whom it was determined.


The most striking finding of this study was the discovery that 21 (51 percent) of these patients died of common medical problems not directly related to tuberculosis. The majority of the patients in this category died of cardiopulmonary disease such as acute myocardial infarction, dysrhythmias, respiratory failure secondary to COPD, cerebrovascular accident, and pulmonary emboli. Since the population studied was primarily elderly, one would expect an increased incidence of underlying arteriosclerotic, cardiovascular, and chronic obstructive pulmonary disease. This undoubtedly was a significant factor in the deaths of these patients. Hypokalemia, which may predispose to fatal cardiac dysrhythmias, has also been recognized as having a possible association with severe tuberculosis; however, electrolyte abnormalities were not found in our patients. The most common single nontuberculous cause of death was pulmonary embolus, which occurred in five of our patients. These patients spent the majority of their time in bed; two had underlying cardiopulmonary disease, and another had nephrotic syndrome. Bacterial infection accounted for three additional deaths.

The patients who died as a direct result of mycobacterial disease demonstrate several pertinent points. Overwhelming disease as a major cause of death occurred in seven patients. Serum albumin levels were uniformly low when determined. No cause for the hypoalbuminemia could be cardiac arrestdetermined. Whether the hypoalbuminemia was a direct result of the tuberculosis or contributed to the severity of the disease (or both) is conjectural. One could speculate that close attention to the nutritional status and correction of deficiencies in patients with tuberculosis might alter morbidity and mortality associated with the disease. Hemoptysis is a well-recognized cause of death and occurred in four of our patients. The presence of bilateral cavitary disease in two or more lobes, as was present in all four patients, precludes surgery; however, Remy et al have demonstrated the effectiveness of transcatheter embolization in the control of bleeding in 41 of 49 patients with massive hemoptysis. Others have also documented the effectiveness of bronchial arterial embolization in controlling massive hemoptysis, which may offer a favorable alternative in the nonsurgical patient Severe respiratory insufficiency was the cause of death in six patients. Interestingly, none of these patients had miliary disease, which was previously the usual cause of respiratory failure in patients with tuberculosis. Our findings are in agreement with the report of Agarwal et al in 1977, who presented data on five deaths in 16 patients with respiratory failure due to fibrocavitary disease. Two patients in our study died as a result of involvement of vital organs. Only one of these patients was suspected of having tuberculosis before death. Clearly, clinicians must be aware of the rapidity with which tuberculous meningitis can progress, the fact that this entity is frequently misdiagnosed, and that prompt diagnosis and chemotherapy is imperative. Clinically undiagnosed pulmonary tuberculosis, frequently with associated miliary disease, has been a common cause of death in several series. Only one of our patients died of miliary disease without a diagnosis of tuberculosis having been made before death.

The finding that seven (35 percent) of those 20 patients dying of tuberculosis and 13 (62 percent) of those 21 patients dying of other causes died within the first month of the institution of chemotherapy is of interest. In none of these patients did it appear that chemotherapy contributed to death, as has been suggested by other authors. The relevance of the delays in the initiation of chemotherapy were difficult to assess in reviewing the records; howevei; they did not seem to play a major role. Seven patients were receiving steroids at the time of death. In only one patient, in whom the diagnosis of tuberculosis had not been entertained, did steroids possibly hasten the patient^ death.

Tuberculosis is still a significant cause of death in this country, even in an era of efletive chemotherapy and supportive measures. In our series, less than half of the patients who died with active tuberculosis died directly as a result of their tuberculosis. Clinicians should be aware that coexisting medical problems, especially cardiopulmonary diseases, bacterial infections, gastrointestinal hemorrhage, and malnutrition play a significant role in the course of patients with tuberculosis. Both prompt recognition and treatment of associated medical problems and a reevaluation of our treatment of the difficult tuberculous problems could decrease mortality.

Table 1—Tuberculous Causes of Death

Mean Age,

yr M/F

Cause of Death No. (Range) Ratio



ength of Thera Prior to Death mo

1-6 7-12








Chest X-R



ay Findings







. of bes >lved


Overwhelming diseases 7 70 (56-89) 7/0 3 2 1 1 7 1/6 4/3 1 6
Massive hemoptysis 4 49 (35-64) 3/1 4 1 0/4 4/0 1 3
Respiratory failure 6 53 (37-79) 6/0 3 1 2 6 1/5 5/1 1 5
Vital organ involvement 2 35 (7-62) 2/0 1* 0 0/0 0/0 0 It
Total drug resistance 1 50 1/0 1 0 0/1 1/0 1