In April 2009, an outbreak of respiratory disease was reported in Mexico, caused by a novel origin swine influenza A (pH1N1) virus. In Chile, located approximately 7000 km south, the Ministry of Health reported the first case of pH1N1 virus on May 16, 2009 (1). The case was a young woman returning from a trip to Dominican Republic. Four days later, the first influenza death was reported in the city of Puerto Montt, 1200 km south of Santiago, suggesting a high capacity of virus spread.
Subsequently, the flu incidence had an increasing trend from south to north in the same way seasonal influenza usually behaves, tending to spread from south to north and paralleling winter cooling in Chile (2). Although a long and narrow country, Chile has a wide range of climates throughout its geography; the south is cooler and wet, whereas the north is dry and hot. The outbreak advanced from Puerto Montt in the south to the central area, and finally to desert areas in the north.
Meanwhile, human-to-human transmission of the virus was documented in at least three countries of two World Health Organization regions. The World Health Organization raised the pandemic level to six on June 11 and declared the first pandemic of the century caused by the spread of new human influenza virus A (pH1N1) (3).
We describe the clinical and epidemiologic characteristics in a case series including the first 75 patients with pneumonia and laboratory-confirmed pH1N1 infection hospitalized at 11 hospitals in Chile. Also, we describe some epidemiologic aspects of the pandemic outbreak in Chile, and we summarize some lessons we learned about the clinical and organizing aspects of this event.
Epidemiologic and Clinical Aspects
The epidemiologic characteristics of all confirmed cases between May 17 and September 23 in Chile were centralized by the Ministry of Health (4). Case criteria were the presence of temperature >38°C, cough, myalgia, headache, and sore throat. Cases were confirmed by a specific real-time reverse-transcription polymerase chain reaction at the Institute of Public Health and some private health centers in Chile.
During the winter, a marked increase of cases of respiratory disease was reported by the epidemiologic surveillance system (Fig. 1). A total of 366,624 cases of pH1N1 were reported through September 23, and 12,248 of those cases were confirmed. The clinical spectrum of the disease caused by new pH1N1 virus infection ranged from a nonfebrile, mild upper-respiratory tract illness to severe or fatal pneumonia. Most cases have been benign and uncomplicated, behaving as typical influenza-like illness with spontaneous recovery after 3 to 5 days.
However, in all patients hospitalized with a diagnosis of suspected severe human influenza infection and/or pneumonia, World Health Organization tests for pH1N1 determination were performed and were confirmed in 1562 cases, with a rate of 9.2 hospitalizations per 100,000 inhabitants. In 132 Chilean deaths, there was a positive World Health Organization confirmation of pH1N1. The death rate was .78 deaths per 100,000 people. In 126 of these cases, influenza was considered the main cause of death; in the remaining cases, influenza was not considered to have a causal role in the death.
Recently, we performed a retrospective analysis (5) of all patients admitted to a critical care beds of 11 national health centers (National Institute of Thorax, Hospital del Salvador, Hospital de Valdivia, Hospital de Coquimbo, Hospital San Borja-Arriaran, Hospital Militar de Santiago, Hospital de la Fuerza Aérea, Hospital de Puerto Montt, Hospital de Osorno, Clínica Indisa, and Clínica Santa Maria), between June 1 and August 30, 2009. In all cases the diagnosis was confirmed by World Health Organization for human pH1N1 virus. We evaluated demographic, clinical, and laboratory findings, number of patients using mechanical ventilation (MV), prone position, high-frequency oscillatory ventilation (HFOV), and extracorporeal membrane oxygenation (ECMO), as well as mortality in the intensive care unit.
We analyzed a total of 75 adult patients admitted to 11 ICU in 11 hospitals in Chile. The median age was 45 yrs (range, 16–77 yrs). Forty-five patients (59%) were male. All patients were Chilean residents. Forty-four had preexisting medical conditions, including obesity (median body mass index, 38 ± 26 kg/m2) in 34 patients (45%), arterial hypertension in 12 patients (16%), asthma in 13 (18%), chronic obstructive pulmonary disease in 10 (13%), non-type 1 diabetes mellitus in three, and chronic congestive heart failure in two. Only eight of the patients had undergone seasonal influenza vaccination in 2008 to 2009; seven of them survived. Our series included seven pregnant women, including one who entered labor because of the seriousness of the respiratory failure (5).
The median time elapsed between the onset of symptoms and admission to the hospital was 5 days (range, 1–15 days). The main symptoms were cough, fever, with temperatures >38°C (100.4°F), dyspnea, and myalgias (Table 1). In this series, 75% of the admitted patients required MV; in four cases, noninvasive ventilation was sufficient for treatment. Prone position ventilation, HFO, and ECMO were used in 18 (24%), eight (10.6%), and five (6%) cases, respectively. ICU mortality was 26%. Survival was seen in 75% of those treated with HFO and in 40% of those treated by ECMO. At admission, 27% (21 patients) had acute renal failure, seven of whom required hemodialysis. Ten (50%) patients who had acute renal failure died. The gastrointestinal symptoms (nausea, vomiting, and/or diarrhea) occurred in <11% of patients (5), which is less than the incidence of 38% reported in the US (6).
The median Acute Physiology and Chronic Health Evaluation II (APACHE II) score was 14 (range, 1–35), and the median Sequential Organ Failure Assessment (SOFA) score was 5 (range, 0–15); both were higher in those who died, indicating more severe abnormalities in those patients than in those who survived (Table 2) (5).
At the time of admission, all 66 (88%) tested patients had elevated lactate dehydrogenase levels. Fifteen cases (20%) had increased creatine kinase levels, exceeding 1000 IU/L (range, 1065–7582) in eight cases (10.6%). In 27 (36%) there was leukocytosis >10,000/mm3, and 21 patients (28%) had thrombocytopenia at admission. Fourteen cases (18.6%) had elevated creatinine levels (1.5–8.5 mg/dL) (5).
Bacterial coinfections were documented in seven cases on admission (four had Streptococcus pneumoniae and three had Staphylococcus aureus). In five patients, pleural empyema developed, and the etiology was identified (bacteremic penicillin-sensitive S. pneumoniae) in one patient (5).
All patients except one had radiologically confirmed pneumonia showing bilateral patchy alveolar opacities (predominantly basal). Interstitial opacities affected three of the four lung quadrants in 45 patients. A rapid progression to ARDS was common, also (5).
We present the management of the influenza pandemic that took place in our country as well as the lessons learned. At the beginning of the H1N1 pandemic, lack of hospital beds for severe pH1N1 cases was identified. In Chile we have 185 hospitals, of which 61 are high-complexity units, and the other 124 are intermediate-complexity and low-complexity units. This issue was managed by means of several adaptive measures (7).
Centralized administration of critical care beds in the metropolitan area was handled on a telephonic and computerized platform called Health Responds. The system tracked vacant basic, acute, and critical care beds, public or private, and consequently could recommend the referral of patients to an available health center's bed. This system complied with nearly 6000 phone requests each day for that purpose during the pandemic (Fig. 2). From May 11 through August 8, 2009, 987 referrals coordinated by this central bed management system were performed; 516 were referred to the public health system and 471 were referred to the private health system.
The emergency guard system was also reinforced with extra personnel, especially in the evening hours. This staff was provided by triage policies designed for determining the management of the patients with different degrees of severity. The elective outpatient care system's schedule was extended to evening hours and weekend days.
Basic acute care beds, mainly from the surgery and gynecology departments, were converted into respiratory care beds, along with an interruption of elective general surgery, orthopedic, and gynecologic procedures. This mechanism allowed the conversion of 50% of extra beds, with an increase from 4261 to 6481 beds for acute disease. In the same way, 25% of the intermediate care beds were converted into intensive care beds (Table 2). Finally, a rate of one intensive care bed for every 20,000 inhabitants was achieved for adult care, and one bed for every 30,000 inhabitants was achieved for pediatric care. At the same time, 20% of the professional staff was absent from work because of respiratory disease, with most of the cases being mild and self-limited. No case of severe pneumonia was reported among them. This whole scenario required reassignment and hiring of extra nursing, physical therapy, and medical staff for the care of pH1N1 patients. All of them were monitored by experienced critical care staff.
As for medical education, conceptual support, and guidelines, the Health Ministry assigned a group of respiratory, intensive care, and infectious disease specialists entrusted to compile the guidelines for management of pH1N1 cases, which were available through the Internet for users around the country (8). Daily videoconferences were held by experts in respiratory and intensive care, leading daily analysis and discussion of cases countrywide. Also, we had tutorial and support activities distributed for the needs in the provinces, by direct visit, mail, or telephone, to assist in sharing and advising in decision-making. Private system, military, and university hospitals and their professional staffs were settled “online” as backup of the public system using the same criteria.
Critical care beds and ventilators were main limiting resources for adequate management of our patients, so 150 new MV devices were purchased by the public system and were distributed throughout the country (Table 3). This number was limited by the number of devices available for sale in the country during the most severe phase of the pandemic. According to the same needs and policies, the private system also purchased equipment. The exact quantity is unknown by the authors.
Advanced critical support for acute lung injury–ARDS patients was resumed. Eleven of the 150 new ventilators were HFOV machines, which were distributed throughout the country according to the technical complexity of each center and the complexity of care needs. Eventually, 17 HFOV machines were available. ECMO was available for 10 patients at a time.
Interhospital transport of critically ill cases, with trips up to 1500 km, were required for severe cases needing advanced modes of therapy, mainly respiratory, surgical, and renal support (9–12). These transports were not available in small and peripheral hospitals. We noticed that many of these cases suddenly became severe, making them ineligible for management with transport ventilators. We considered patients eligible for air medical transport provided they had an oxygen saturation of ≥90% while undergoing testing on transport ventilator in volume- or pressure-control modes, an inspired oxygen fraction up to 1, positive end-expiratory pressure ≤15 cm of water, eventual pneumothoraces were drained, and were hemodynamically stable with two or less vasoactive drugs. The transport of patients needing HFOV or ECMO used to be technically impossible, even after hours working at the bedside with the ventilatory, sedation, and relaxation parameters. Our recommendation is to consider the early referral of the most severe cases, because that referral may not be possible later in the evolution.
Instructions were disseminated to provide antivirals widely, free from cost, to the cases properly diagnosed, certified, and registered by physicians. Nearly 1,000,000 doses were acquired (and 662,428 treatments administered by September 20) and distributed to outpatient and emergency care systems, including private, university, and military institutions.
Lessons Learned About the Clinical Management of Severe Pneumonia in Human Infection With New Influenza A Virus: Advanced Respiratory Support
Treatment of ARDS associated with the pH1N1 virus infection should be based on published evidence-based guidelines for sepsis-associated ARDS. Lung-protective MV strategies were used (13). Although noninvasive ventilation was discouraged because of the risk of spreading viruses, some centers made limited use of noninvasive ventilation, with clinical success in properly isolated cases, and with the staff using adequate means for protection. When invasive ventilators with noninvasive ventilation software and closed masks were used, these limitations were considered to be not that restrictive.
The 17% of patients with severe respiratory failure supported by HFOV or ECMO needed prolonged support, up to 21 days in our series. Whereas this course seemed disappointing initially, eventually we learned that it did not necessarily predict a poor outcome.
Despite the initial improvement of oxygenation in our series, three in five patients (60%) died and one survived with severe sequelae. We believe that implementation of this complex technique should be reserved for well-trained and experienced centers.
HFOV had a lower mortality in our series (two of eight patients) and, despite being virtually unknown technology in our country, it could be implemented quickly in different regions. We reserve it as an early rescue strategy for patients with oxygenation index >15 or Fio2 .7 with positive end-expiratory pressure >15 cm H2O.
On admission, a restrictive fluid strategy was attempted (14) for the management of patients with acute respiratory failure secondary to pH1N1, provided they were not experiencing hypoperfusion or a state of shock, before the connection to MV. Centers provided with PiCCO technology (Pulsion Medical Systems, Munich, Germany) to measure extravascular lung water reported it was increased from the onset of respiratory failure, reflecting impaired permeability. Once respiratory failure and MV are established, the strategy consists of supplying the least volume input compatible with maintaining hemodynamics and renal perfusion. Care must be taken in the interpretation of elevated pulmonary artery catheter figures in patients requiring high airway pressures.
The pH1N1 in our country, located in the southern hemisphere, was associated with winter, generating an increase in demand for medical care and numerous hospitalizations for severe cases. This situation necessitated the development of strategies to optimize the use of scarce medical resources in the highly demanding time of winter. We communicate the lessons that we learned about the clinical presentation and treatment of a new form of influenza.