A 70 year old retired police inspector presented to the emergency department with a 10 day history of fever followed by rapidly progressive dyspnoea for 3 days prior to admission. He was a non-smoker and teetotaller and there was no prior history of any medical illness. Physical examination revealed marked tachypnoea (respiratory rate 40 breaths/min), central cyanosis, and extensive fine inspiratory crackles on auscultation of the chest. Haemogram revealed haemoglobin 124 g/l and total leukocyte count 30 000/mm³ with a differential of 84% neutrophils, 9% lymphocytes, 5% monocytes, and 2% eosinophils. Erythrocyte sedimentation rate was 35 mm/hour and platelet count 263 000/mm³. Renal and liver function tests were within normal limits. There was severe type 1 respiratory failure on arterial blood gas analysis (table 1). Chest x ray revealed bilateral diffuse alveolar infiltrates with normal cardiac size (fig 1A). The patient therefore fulfilled three of the four criteria for the diagnosis of ARDS.¹ Extensive ground glassing was observed on high resolution computed tomography (CT) scan of the chest (fig 1B). Serum cold agglutinins were positive, with a titre of 1:64. Blood culture was sterile and there was no evidence of deep venous thrombosis on compression ultrasound of the lower limbs.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

 Left panel, bilateral alveolar infiltrates on chest x ray with normal cardiac size consistent with ARDS; right panel, a series of high resolution chest CT scans demonstrating bilateral extensive ground glass opacities.

Broad spectrum antibiotics (intravenous cefotaxime and levofloxacin) and high flow oxygen (FiO₂ 0.5) were administered with no significant improvement in either clinical condition or blood gases over the next 48 hours (table 1). The patient was shifted to the respiratory intensive care unit (RICU) of the Institute, anticipating the need for mechanical ventilation. In the RICU, he was put on NIPPV (VPAPII, Resmed) with a nasal mask in addition to the treatment already being given. Initial settings were inspiratory positive airway pressure (IPAP) of 8 cm H₂O and expiratory positive airway pressure (EPAP) of 4 cm H₂O. Within an hour of starting ventilation there was marked improvement in respiratory rate, which fell from 40 to 22 breaths/min, dyspnoea decreased, and a significant improvement in blood gases was noted (table 1). This improvement was sustained and progressive except for episodes of severe desaturation when removal of the mask was required for eating and drinking. The maximum pressures administered during his hospital stay were IPAP 12 and EPAP 8 cm H₂O. Levels of IPAP and EPAP were gradually reduced as the patient’s clinical condition and blood gases showed a steady improvement. He was discharged after a hospital stay of 11 days with a satisfactory arterial blood gas result in room air, and advised to continue levofloxacin for a period of 3 weeks. On follow up a month later, the patient was asymptomatic, his cold agglutinin titres had declined to 1:8, and there was almost complete clearing on chest x ray.

DISCUSSION

It is now recognised that Mycoplasma pneumoniae, traditionally believed to cause mild disease, is the aetiological agent in about 2–7% cases of severe community acquired pneumonia (CAP).^2,3 However, ARDS due to M. pneumoniae is distinctly unusual, with only sporadic case reports in the English literature.^4,5

NIPPV has now assumed a central role in the management of hypercapnic respiratory failure due to chronic obstructive pulmonary disease (COPD) and in cardiogenic pulmonary oedema; however, its role in ARDS secondary to severe pneumonia is highly controversial. Early studies comprising a heterogenous group of patients with acute respiratory failure treated with NIPPV suggested that pneumonia was a poor prognostic factor. As an illustration, in Antonelli’s study,⁶ all patients with pneumonia randomised to NIPPV failed therapy, whereas there was a 60% success rate in those patients in whom respiratory failure was not attributable to pneumonia. The authors speculated that the higher failure rate in patients with pneumonia was due to difficulty clearing secretions, reduced pulmonary compliance, and nonhomogeneous gas exchange. Analysis of the first randomised trial of NIPPV in 56 patients with pneumonia revealed that only the subgroup of patients who had concomitant COPD appeared to benefit from NIPPV.⁷

In a study by Rocker et al,⁸ there was a 66% success rate when NPPV was used as the initial mode of assisted ventilation in 10 patients with ALI/ARDS. More recent data exclusively from non-COPD patients with severe CAP revealed that 22 of 24 patients had initial improvements in oxygenation and respiratory rates after starting NIPPV but that 66% of the study patients eventually required intubation.⁹ The authors found that patients whose respiratory rates decreased at the end of the first hour of NIPPV were less likely to require intubation than those in whom the respiratory rate either remained stable or increased. Our patient had shown a significant decrease in respiratory rate as well as dyspnoea at the end of the first hour of NIPPV. However, we feel that this dictum of clinical improvement at 1 hour predicting a successful outcome of NIPPV is applicable more to cases of COPD than severe pneumonia, as the latter is characterised by alveolar filling, ventilation perfusion mismatch, and decreased pulmonary compliance, all of which are less rapidly reversible than the hypercapnic respiratory failure of COPD, in which respiratory muscle fatigue plays a predominant role. The most recent and largest study (105 patients) to address the issue of NIV in severe pneumonia found that the use of non-invasive ventilation compared with high concentration oxygen therapy decreased the need for intubation, the incidence of septic shock, and intensive care unit mortality, even though the seven patients with ARDS had a poor outcome (five died).¹⁰ Current evidence therefore suggests, as was demonstrated in our case, that NIPPV is warranted in most patients of ARDS due to severe community acquired pneumonia not responding to standard medical therapy. However, as intubation rates are high, NIPPV should ideally be used in such patients with a conventional mechanical ventilator on standby. It is the subgroup of patients with relatively mild, early ARDS who are likely to benefit from NIPPV and avoid intubation.