Without taking a detailed history, you may be missing a serious diagnosis. For example, as a case vignette, an elderly lady was admitted to a hospital with worsening dyspnoea. She had a background of end-stage COPD from 80 pack-years (Brinkman 1600) of smoking.
Unfortunately, the history was incomplete (and therefore so was the subsequent assessment and plan) because the resident had failed to elicit from the patient that she had an active cough and sputum.
The chest radiograph revealed bilateral upper lobe infiltrates, a simple pneumothorax and hyperinflation consistent with COPD.
The patient was admitted for 'observation' to an open ward with other patients and was not wearing a face mask.
Examination revealed an emaciated patient with a hyper-expanded chest. She was mildly febrile (37.5 deg C), respiratory rate 24/min, HR 90 regular, BP 130/80mmHg and O2 Sats 98% on 2L oxygen via nasal cannulae. Lymph nodes were not palpable. There was increased percussion sound throughout the chest but reduced air entry in all lung fields and crackles at the left apex.
The rest of the examination was within normal limits.
In view of the high prevalence of tuberculosis in Japan, this should be the very first consideration in any patient with upper lobe infiltrations with pneumothorax. Although COPD is the commonest cause of pneumonthorax (70%), the concommittant existence of both radiographic features point primarily towards Tb. At the bedside, more history was taken and it was revealed that the patient did indeed have a cough and sputum. In fact, she coughed up sputum right in front of the medical team.
There are several other causes of upper lobe infiltrates defined by the following mneumonic:
S - Sarcoidosis, Sulfa drugs, Silicosis
H - Histoplasmosis
I - Idiopathic
R - Radiation induced
T - Tuberculosis (post-primary)
C - Chronic allergic alveolitis
A - Ankylosing spondylitis, Aspiration penumonia
P - PJP (when using pentamidine nebulisation), Progressive Massive Fibrosis
E - Eosinophilic granuloma
On closer inspection of the chest radiograph, the patient had vertebral body changes with calcified ligaments that might be consistent with ankylosing spondylitis. However, sometimes, severe facet joint osteoarthritis can present in this manner in the elderly as can Diffuse Idiopathic Skeletal Hyperostosis. Ankylosing spondylitis usually presents in a younger age group and is relatively rare. Osteoarthritis is very common and a more likely explanation for the 'bamboo spine' changes.
However, given the above scenario, what is the likely cause for bilateral upper lobe infiltration in a patient with COPD?
Of course, atypical presentation of a community acquired pneumonia is possible. e.g. Streptoccous pneumoniae. Moreover, Klebsiella pneumonia, H. influenza, Pseudomonas or even post-influenzal staphylococcal infection can present in such a way. Atypical pneumonia should also be considered. Of course, this patient might have aspiration but bilateral infiltration with the absence of lower lobe infiltration would make this diagnosis somewhat tenuous.
However, the above infections can be diagnosed rapidly and treated quickly and effectively with standard antimicrobial agents. Tuberculosis on the otherhand is difficult to diagnose and requires multiple antibiotics for many months. Moreover, Tb is very infectious and has an infectivity rate of about 10% ! That translates to 1 in 10 persons becoming infected and developing active disease at some point during their life from a patient coughing up active Tb into a space where the Tubercle bacilli remain airborne.
Hence, anyone with even the remotest suspicion of active tuberculosis who is coughing up sputum must not be placed on an open ward. They must be admitted directly to an isolation room with negative pressure (air is sucked into the room). Full protective garments should be worn by the medical staff including the 3M filtration face masks.
Such patients should be kept isolated until such time that tuberculosis has been excluded. Please see any good infectious disease book for the work-up of tuberculosis e.g. Dr Makoto Aoki's Infectious disease book.
It is no good trying to make an simple diagnosis e.g. aspiration pneumonia, without first excluding the serious/life threatening diagnoses that one 'should not miss'.
As I have heard many times in Japan 'In any patient with a fever, always consider tuberculosis'. I think that is a very apt saying and which I fully support.
Further widely quoted advice is 'rule out Tb before starting the steroids'. Again this is very important. Corticosteroids can reactivate latent Tb or worsen existing Tb. So be careful to check the PPD skin test, sputum examination and culture before pulsing your suspected interstitial pneumonia patient with methylprednisolone which might actually be Tb.
Usually, interstitial pneumonia patients can afford to wait for several days before starting the steroids giving enough time for the preliminary Tb examinations to be done. However, if steroid therapy is necessary because of worsening disease or for example, the patient has an 'infective exacerbation of COPD' requiring steroids with a suspicious CXR suggesting Tb, anti-Tb therapy should be commenced with the steroids (preferably after Tb samples have been taken) in order to treat this suspected infection along with that of the the acute illness. If Tb is later ruled out by tests, the Tb therapy can be stopped. It is better to be safe than sorry.
This is supported by UpToDate 17.1 in the section under Major Side Effects of Systemic Glucocorticoids which says 'Conventional wisdom suggests that any patient with a positive PPD skin test, a suggestive chest radiograph, or a strong family history of tuberculosis should receive antituberculosis prophylaxis prior to, or concomitantly with, glucocorticoid therapy'.
You cannot take the risk of having a single patient on an open ward with active Tb. Yes, it may cause the nursing staff / doctors / managers some extra work in order to organise an isolation room, but by not doing so puts many people at risk of Tb exposure.