Samples should be collected once the suspected clinical diagnosis of influenza infection has been established in patients requiring hospitalization or admission to the ICU. Respiratory samples should be obtained as soon as possible after the onset of symptoms (ideally in the first 48–72h), in order to maximize the sensitivity of the techniques. In hospitalized patients, the samples can be collected from different points of the upper (UA) or lower airway (LA). In ventilated patients we should obtain respiratory samples from the upper tract, in the form of smears or nasopharyngeal aspirates, and also from the lower tract (bronchial aspirate [BAS], bronchoalveolar lavage [BAL] or miniBAL). Viruses can be detected in LA samples in up to 20% (1 out of every 6 in the Spanish series) of all patients with viral pneumonia who present negative UA samples.8,9 Consequently, BAS or BAL should be performed whenever possible in those seriously ill patients with suspected viral pneumonia10,11 and, if the tests prove negative, they should be repeated in the subsequent 48–72h. Occasionally, some patients require more than three samples to yield a positive result.9

Identification of the viral elimination period can help us establish the duration of the adopted infection control measures, thereby avoiding potential in-hospital outbreaks,12 particularly in the ICU, where such problems can prove fatal. Infection isolation and control measures are essential for controlling the transmission of influenza, and must be continued until the diagnostic test results prove negative (i.e., until negative conversion). The influenza virus persists in the secretions of immunocompetent patients for 5–7 days on average, though the elimination or clearance period may be prolonged particularly in elderly people, small children, patients with chronic diseases, immunocompromised individuals, more seriously ill patients, and in those who develop acute respiratory distress syndrome (ARDS).13,14

The early use of oseltamivir can shorten the viral clearance period.15 We therefore consider it necessary to collect new respiratory samples from 7 to 10 days after the start of symptoms, to confirm RT-PCR test negativity and be able to suspend the infection isolation and control measures—facilitating the vigilance and care of critical patients subjected to mechanical ventilation in isolation wards. No studies have assessed the risk of influenza transmission in those patients in which the PCR test results or culture findings remain positive during prolonged periods of time, though in theory these individuals would be able to transmit the infection. Based on these premises, individual evaluations will be needed in order to define the duration of the isolation measures.

The diagnosis of influenza A/H1N1 is fundamentally clinical, particularly during the seasonal influenza peaks or in epidemic outbreaks. Diagnostic tests are advised in all patients with fever and flu symptoms who require admission to hospital. Respiratory samples (nasal or pharyngeal exudate or deep samples in intubated patients) are to be obtained as soon as possible, with the immediate start of empirical antiviral treatment.

In Spain, the risk factors associated to the development of serious disease and to admission to the ICU in the recent influenza A/H1N1 pandemic were respiratory disorders and morbid obesity, while the most frequently documented complications were respiratory distress related to the development of primary viral pneumonia, hemodynamic instability, acute renal failure and multiorgan failure. The mortality rate among the patients admitted to Spanish ICUs was 22–25%, with a global mortality rate of 0.43 deaths per 1000cases.

• (a)

  The rapid and early detection of infection caused by the new pandemic virus, particularly in patients with criteria for admission to the ICU, allows us to optimize treatment of the disease, and is decisive in the adoption of effective epidemiological control and prevention measures.50,51

• (b)

  The molecular methods based on nucleic acid amplification techniques have become the gold standard in the diagnosis of the influenza A/H1N1 infection, at the expense of classical cell cultures,50,51 and have been included by the United States Centers for Disease Control (CDC) in the definition of “confirmed” cases of the disease. The extreme sensitivity and specificity of these methods allow an early, precise and rapid diagnosis (within a few hours) of the infection. A range of RT-PCR protocols have been developed and clinically evaluated, and some of them have been marketed and approved by the United States Food and Drug Administration (FDA) for diagnostic use. These tests detect genic sequences specific of the pandemic virus and pertaining to the genes that encode for hemagglutinin (most tests), neuraminidase or M protein. Likewise, PCR multiplex procedures have been developed in which the targets are detected by means of hybridization techniques (ResPlex II, Luminex X-TAG, among others), and which allow us to simultaneously examine the presence of multiple respiratory pathogens, as well as several subtypes of influenza viruses.52

• (c)

  The immunochromatographic methods that were commercially available before eclosion of the pandemic virus generally lack sensitivity in diagnosing the new influenza A/H1N1 virus (11–80%), and do not allow us to differentiate among the different subtypes of the virus.53 The analytical limit of detection of these tests is approximately 1.0×105TCID50/ml.54–56 Their sensitivity is critically dependent on the type and quality of the sample (number of epithelial cells present), the age of the patient (being more useful in children than in adults, particularly when sampling is performed in the first 24–48h after infection, since children eliminate a larger number of viral particles from the upper airway), the delay in obtaining the sample after infection, the rapidity of sample transport, and the care with which the sample is processed.53 Consequently, a negative result of these tests does not rule out the possibility of active infection. New immunochromatographic methods have been developed and marketed that allow the specific diagnosis of infection caused by the pandemic virus, offering a sensitivity of about 30% in nasopharyngeal exudates, and with a limit of detection of between 2.6×104 and 1.0×105TCID50/ml57,58—though they generally have a high positive predictive value.

• (d)

  Classical culture in MDCK cells is slow and of low sensitivity. The appearance of cytopathic effects may take 2–14 days. The combined use of a shell vial method (with A549 cells and Mv1Lu cells) and a pool of monoclonal antibodies against different subtypes of influenza viruses (and other respiratory viruses) offers a diagnosis within 24–48h, with a sensitivity comparable to that of classical culture.59,60

• (e)

  Direct immunofluorescence with samples obtained from the UA or LA is an adequate alternative to the immunochromatographic tests, with the added advantage of allowing us to assess the quality of the sample obtained. Its sensitivity is approximately 90% compared with the molecular methods when the sample is of optimum quality (>30 columnar epithelial cells per well) and the slides are prepared in centrifuge, allowing the exclusion of inadequate preparations.59,60

• (f)

  The indirect specific antibody detection methods, which allow us to establish a confirmative diagnosis after initially elevated levels or the confirmation of seroconversion against the pandemic virus, are performed with the different patient serum samples, which are submitted and stored for processing in parallel—the true interest of these techniques being confined to epidemiological studies.

The viral load in the UA reaches its peak during the first two days after onset of the symptoms, and quickly decreases over time, in direct relation to the increase in local and systemic neutralizing antibodies, which can be quantified using seroneutralization procedures. In uncomplicated cases we usually do not detect viral RNA in the UA at the end of the first week after symptoms onset.50,51 However, children and immune depressed patients can excrete infective virus for a longer period of time. Taking these variables into account, respiratory sampling should be carried out as soon as possible after appearance of the symptoms, without excluding samples at later timepoints, depending on the type of patient involved (e.g., complicated cases with severe respiratory symptoms in patients admitted to the ICU).

Nasopharyngeal aspirate offers a greater diagnostic yield than nasopharyngeal, oropharyngeal or nasal exudate,50,51 though a recent study has shown that the combined use of nasal and pharyngeal exudates affords the same diagnostic yield as nasopharyngeal aspirate.61 According to the recommended method, nylon flocked swabs are preferable to conventional swabs, since they are more effective in collecting epithelial cells.62 In the case of intubated patients in the ICU, it may be useful to process specimens from the LA (due to the greater viral load involved),27 in all cases taking the necessary precautions against aerosols in BAL or BAS. Even in the event of impossible intubation due to the severity of the patient condition in admissions to the ICU we could combine the initial nasopharyngeal exudate or aspirate yielding negative results with the processing of tracheal aspirate samples, thereby establishing a confirmatory diagnosis.63 However, the diagnostic efficacy of the molecular and immunochromatographic tests applied to these LA samples has not been sufficiently evaluated to date.

In the event of clinical suspicion, the collection of respiratory samples for establishing an etiological diagnosis of influenza A/H1N1 should be carried out following the methodology described above—it being generally advisable to limit sampling and microbiological processing to those cases of infection caused by the pandemic virus in patients who require hospitalization. The necessary diagnostic tests should be made in the Department of Microbiology, and the ultimate choice of test depends both on the characteristics and intervention possibilities defined by the type of patient and/or sample involved, and on the availability of the different techniques in each hospital center. Where available, the RT-PCR molecular methods are the recommended first choice, and the results should be available within 24h after sample reception. Due to its rapidity, high positive predictive value and easy performance, an alternative first choice could be an immunochromatographic technique capable of specifically detecting the pandemic virus. However, if this option is chosen, we must remember that because of its lesser sensitivity, a negative result in this case does not necessarily imply the absence of active infection. Consequently, the samples must be kept under conditions suitable for processing with the molecular nucleic acid amplification techniques, which can ultimately confirm the diagnosis. Clearly, in the case of initially negative microbiological test results (including RT-PCR) in seriously ill patients admitted to the ICU with strong clinical suspicion, we should attempt to establish an etiological diagnosis through processing of the samples, including LA samples where possible. In any case, the clinical picture and patient course will ultimately guide the management decisions.27,63

Lastly, in 2009 the prevalence of influenza A/H1N1 strains resistant to oseltamivir64 (H275Y mutation) was less than 1%. Genotyping techniques for the detection of resistance to oseltamivir are to be preferred over functional viral neuraminidase inhibition (phenotypical) methods, which are only available in specialized centers. Accordingly, it is not advisable to include resistance studies in the routine microbiological diagnostic protocol applied to processes of this kind.

Molecular methods based on nucleic acid amplification techniques (RT-PCR) are the gold standard for diagnosing influenza A/H1N1 infection. The immunochromatographic techniques are of limited sensitivity, as a result of which negative readings do not discard the possibility of active infection. Classical cell cultures both lack sensitivity and are slow. Direct immunofluorescence offers a sensitivity of 90%, but requires a good quality sample. The indirect antibody detection methods are only of epidemiological interest.

From the existing data we can conclude that although patients with influenza A/H1N1 infection can present relative leukopenia, with elevations of LDH, CPK and PCR, these laboratory test parameters are not independently related to the patient prognosis. However, plasma LDH>1500U/L and the presence of thrombocytopenia could define a population of patients at risk of suffering serious complications.

Different health organizations4,10,81 coincide that antiviral treatment should be started as soon as possible in all “possible” or “confirmed” cases of influenza A/H1N1 requiring hospitalization, in patients with severe progressive disease, or in the presence of complications, and that this should be done independently of the previous patient health condition or antecedents of vaccination. However, in Spain the mean time to administration of the antiviral agent from the onset of symptoms was 4.3 days during the influenza A/H1N1 pandemic,82 with no differences between the patients who survived (4.1 days) and those who died (4 days). This time interval was a little shorter than that reported by Domínguez-Cherit et al.44 in Mexico (6 days), but longer than that recorded by Jain et al.83 in the United States (3 days) and by Cao et al.71 in China (1 day). Surprisingly, during the influenza epidemic of the winter of 2010, the mean delay in starting antiviral therapy was even longer (5 days on average).37

According to the available data referred to the epidemiological vigilance and control of resistances in recent years, oseltamivir and zanamivir remain the antiviral drugs recommended for the treatment of influenza A/H1N1 infection, since over 99% of the currently circulating viruses are sensitive to these agents.84,85 Amantadine and rimantadine are other antivirals that have been associated to treatment with oseltamivir during the pandemic but should not be used, due to the high resistance rates against these agents found among the circulating influenza A/H1N1 strains. Oseltamivir is a neuraminidase (NA) inhibitor which after absorption in the gastrointestinal tract is quickly converted to its active form, oseltamivir carboxylate (OC), with a bioavailability of over 80% and important penetration of the lung tissues.86 The recommended adults dose is 75mg twice a day, via the oral or enteral route. Since OC is mainly excreted through the kidneys, the dosage must be adjusted to the renal function of the patient. In practical terms, in patients with a creatinine clearance of between 10 and 30ml/min, the dose should be only 75mg/day. There is limited experience with the effectiveness of oseltamivir administered via the enteral route in critical patients, which may suffer alterations in gastric motility, and lesser intestinal absorption of the drug. This situation in turn can be worsened by the increase in distribution volume observed in such patients, particularly when mechanical ventilation is required—giving rise to higher plasma drug levels and therefore to lesser antiviral effectiveness. Some authors87,88 have suggested the need to maintain high plasma concentrations or to secure areas under the concentration–time curve (AUC) equivalent to over 50% of the maximum inhibitory concentration (MIC) for the influenza virus, in order to ensure optimum suppression of viral replication. Following the recommendations of the WHO,89 over 70% of the patients admitted to Spanish ICUs during the influenza A/H1N1 pandemic received high-dose oseltamivir (300mg/day). However, the administration of higher oseltamivir doses was not associated to lesser mortality82 (Fig. 2A and B). These findings are consistent with those of a recent study that analyzed the enteric absorption of oseltamivir in critical patients.90 The authors found that the usual doses of 150mg/day resulted in similar plasma concentrations in both critical patients and outpatients. They therefore concluded that the standard dose is sufficient for reaching the desired concentrations in critical patients. However, some patients showing a lack of treatment response to oseltamivir or who cannot receive the medication via the enteral route because of intolerance can be candidates for treatment with an alternative antiviral agent. In this context, zanamivir is the other NA inhibitor that has been approved for the treatment of influenza A/H1N1. In contrast to oseltamivir, it is not absorbed via the enteral route, and therefore must be administered as a powder for inhalation. While suitable for patients with less serious disease, this drug formulation cannot be used in ventilated critical patients. Only 10–20% of the inhaled dose is absorbed, and the extrapulmonary distribution of the drug is minimal.86 The adult dose is 10mg every 12h, and although there are no clear data, it should be lowered in patients with severely impaired kidney function. Zanamivir cannot be nebulized in intubated patients, since it can cause obstruction of the respirator circuits, severe bronchospasm and death.91 Zanamivir via the intravenous route has been suggested as an alternative in patients who fail to respond to oseltamivir or who have developed resistance. Because of the medical emergency posed by the influenza A/H1N1 virus, this formulation has received special treatment, and has been used on a compassionate basis. Few conclusions can be drawn from the limited published cases, since in all instances the drug was used as rescue therapy. A recent report on 6216 cases documented and hospitalized during the pandemic in the United States found only 8patients to have received zanamivir via the intravenous route. In Spain, the use of zanamivir via the intravenous route was anecdotal during the pandemic (0.7%), but increased to 6.4% (33 patients) during the post-pandemic epidemic of 2010. This increase in use does not imply the appearance of resistances during this phase; rather, it reflects the data generated by a safety study on zanamivir conducted during that period. Until further information becomes available, zanamivir via the intravenous route would only be indicated in patients with strongly suspected or confirmed influenza A/H1N1 infection resistant to oseltamivir. Lastly, peramivir, an investigational antiviral drug that received exceptional approval during the pandemic period, has been used with apparently good results in the United States92 and China.93 Developed for intravenous administration, this drug can be used in critical patients without the potential complications of the enteral route. A daily dose of 300–600mg has been found to be safe and lessens the symptoms of the disease, though its impact upon mortality is still difficult to define. Theoretically, and if supported by new results, this would be the antiviral drug of choice in critical patients, due to the possibility of intravenous administration.

Figure 2.

Mortality in the ICU in relation to the administered oseltamivir dose and severity as assessed by the APACHE II score in 661 patients during the 2009 pandemic (A) and in 430 patients in the post-pandemic period (B). ICU: Intensive Care Unit.

(0.16MB).

The duration of antiviral treatment in critical patients is another controversial issue. In patients with mild or moderate disease, the recommended duration is no more than 5 days.4,81,84,85 This is related to the time to viral clearance from the airways (about 4–5 days from onset of the symptoms). However, in immune depressed patients or subjects with major comorbidities, as well as in critical patients, this viral elimination period can be significantly extended. Leekha et al.94 found over one half (54%) of the patients with influenza A to remain positive for the virus after 7 days of treatment, as determined by PCR. No studies allow us to conclude when to suspend antiviral treatment in seriously ill patients, though in the Spanish ICUs the patients received a median of 10 days of treatment.44,71,82,83,95 Recent studies have described much longer PCR positivity for the virus in critical patients, particularly in those with pneumonia. Ling et al.15 found the mean elimination time to be 6 days, though the virus was still detected in 37% of the patients by day 7, with no apparent relationship to early therapy. In turn, Meschi et al.96 found the presence of pneumonia to be associated with a longer viral clearance time. The authors showed PCR testing to remain positive in 100% of the patients with pneumonia on day 9 of treatment, and in 67% of the patients after 10 days. Although the persistence of viral elimination does not necessarily imply infectious activity, from the health safety perspective we suggest extending treatment at least 10 days in critical patients who evolve favorably, or until negative conversion of the PCR findings. In the case of slowly evolving cases with persistently positive PCR findings, a respiratory sample culture should be obtained to confirm non-viability of the virus, with the suspension or not of treatment, and also to discard the appearance of resistance during treatment.

As has been mentioned, few studies have examined the impact of antiviral treatment upon patient mortality during the pandemic of 2009. The use of antiviral agents during the pandemic was based on the cumulative experience gained with seasonal influenza (1970–2008), where early treatment with NA inhibitors (i.e., administered within 48h of symptoms onset) was seen to reduce the severity and duration of the disease and possibly also the risk of complications.79,80 However, in contrast to seasonal influenza, the severe cases of influenza A/H1N1 show a different behavior, not only because this virus affects younger people, but also because it has been shown to be able to produce rapid and serious lung problems – including the development of ARDS and death secondary to refractory respiratory failure. In reference to this special context it is difficult to extrapolate the impact of antiviral treatment from studies in which the patients suffer only mild or moderate influenza infection.

A recent review95 of 11 published studies found that the complications of influenza A/H1N1 can be lessened by the use of antivirals in both high and low risk patients—though the reviewed series involved important methodological limitations. In turn, Jain et al.83 found the only variable independently associated to mortality to be the administration of oseltamivir within the first 48h, while Domínguez-Cherit et al.44 observed that those patients who received antiviral drugs had a greater probability of survival (OR=7.4). Recently, three retrospective studies97–99 have evaluated the impact of treatment during the pandemic. Hiba et al.97 studied 506 hospitalized patients and found early treatment to be associated with a lesser incidence of complications and with lesser in-hospital mortality. However, this study only included a minimum number of critical patients (n=30; 12.4%). In turn, Higuera Iglesias et al.98 have analyzed the clinical course of 442 patients administered with oseltamivir during the pandemic in Mexico. The authors found that both the probability of developing pneumonia and the risk of suffering severe pneumonia increased with the length of delay in administering the antiviral drug. However, only a little over one half of the patients (54.4%) were hospitalized, and no mention was made of how many of these subjects required admission to the ICU. Lastly, Yu et al.,99 in over 1200 hospitalized but non-critical patients, found early treatment to be associated with a lesser development of radiological infiltrates and with a reduction in the viral elimination time.

The GETGAG/SEMICYUC has published the only prospective study examining the impact of early treatment with oseltamivir during the pandemic in a large number of critical patients (n=657).82 This study found early treatment with oseltamivir to imply increased survival, but only in the more seriously ill patients, i.e., those with invasive mechanical ventilation who had received effective antiviral treatment (at least 4 doses). Furthermore, the impact was so notorious that one life could be saved for every 8 patients receiving early treatment with oseltamivir. These data derived from multivariate analysis were also confirmed by the corresponding propensity score, a statistical model that calculates the probability that a patient will receive early or late therapy, with adjustment for the possibility that the observed differences in clinical course are due to differences between groups and not to greater efficacy of early treatment (Table 4). However, the impact of antiviral treatment in less seriously ill patients (e.g., without mechanical ventilation) is difficult to establish, since the mortality rate recorded in these patients is low. Therefore, most studies center their primary objectives or endpoints on the time to resolution of the respiratory symptoms or fever, instead of on mortality. In the post-pandemic period during the winter of 2010–2011, and despite the observation of an increased delay in administration of the antiviral medication,37,74 early treatment with oseltamivir was also associated with an improved outcome and a 50% reduction in the risk of death (hazard ratio (HR)=0.49; 95%CI 0.28–0.87; p=0.01) (Fig. 3) (non-published data).

Figure 3.

Adjusted survival analysis (Cox regression) in 397 critical patients admitted to the Intensive Care Unit in the post-pandemic period according to whether oseltamivir was administered early (≤48h) or late (>48h). The censored model after 60 days showed survival to be significantly greater in patients with early administration of the antiviral agent. ICU: Intensive Care Unit.

(0.08MB).

Lastly, we agree with the CDC (81) that patients with a high suspicion of influenza A/H1N1 should continue with treatment, independently of the negative results of the initial tests, unless an alternative diagnosis can be established or the clinical criteria suggest a low probability of influenza.

Although there is a lack of randomized and controlled studies on the effect of antiviral treatment in critical patients with severe influenza A/H1N1 infection, the experience gained in these last two years allows us to recommend early administration of the antiviral drug (oseltamivir) (<48h from symptoms onset), at a dose of 75mg every 12h and with a duration of at least 7 days, or until clinical improvement becomes evident. The existing antivirals, particularly those formulated for intravenous administration, could prove to be the drugs of choice in future epidemics.

Primary viral pneumonia is defined as the appearance of lung infiltrates (condensations) during the acute phase of influenza infection and affecting two or more lung lobes, with respiratory sample and blood cultures proving negative for bacterial growth. Confirmation of viral pneumonia due to influenza A is assumed when in addition to the above, identification of the influenza A/H1N1 virus is established by RT-PCR or virological culture.

Community-acquired or healthcare-related respiratory coinfection is defined as any bacterial respiratory infection diagnosed in a patient with influenza A/H1N1 within the first 48h of hospitalization. After this time period, bacterial coinfection is taken to be nosocomial.75

Bacterial coinfection has been associated to viral infection. The condition can manifest as simultaneous infection (coinfection) or as a complication secondary to influenza infection, manifesting when the viral infection is already in the resolution phase. Classically, Staphylococcus aureus (S. aureus) has been implicated in bacterial coinfection that tends to take a slow or torpid course, with lung infiltrations showing early cavitation (<72h), and which implies important mortality. In contrast, S. pneumoniae, Haemophilus influenzae (H. influenzae) and other streptococci tend to be responsible for later conditions manifesting in those patients who initially improve of their influenza infection and who after approximately one week experience clinical worsening (return of fever), with the appearance of new condensations on the chest X-rays. In these patients the prognosis tends to be more favorable.101

The incidence of detection of coinfection during the influenza A/H1N1 pandemic has varied among different countries. In Canada and Australia the detection rates were 32% and 20%, respectively, while in Spain the percentage reached 17.5%—this figure being considerably lower than the 28.6% recorded in a necropsy-based study in the United States. However, this study, which included 77 patients, had important design limitations, since the included cases did not correspond to a systematic analysis, and thus the series may not have been representative of patients with influenza A/H1N1 infection. Consequently, the observed frequency of coinfection cannot be taken as a measure of the prevalence of this complication, though it does confirm the participation of coinfection in the patients who have died.27,29,75,100

S. pneumoniae, H. influenzae and S. aureus were the most frequently isolated pathogens in patients with bacterial coinfection. Specifically, in Spain, S. pneumoniae was found to be responsible for coinfection in 54.8% of the cases, followed by methicillin-sensitive S. aureus (8%) and Streptococcus pyogenes (5.3%).

During pandemic situations or in the period of seasonal influenza, patients with flu syndrome and acute respiratory failure presenting lung infiltrations and clinical manifestations of pulmonary infection, and who require hospitalization or admission to the ICU, should receive empirical antibiotic treatment (ATB) concomitant to antiviral therapy. In order for this strategy to minimize ATB overuse, it is necessary to either confirm or rule out the presence of coinfection by obtaining good quality respiratory samples (from deep locations if possible) immediately after orotracheal intubation, and whenever blood cultures and urinary antigen tests for S. pneumoniae and Legionella spp. are made.

There are no clinical or epidemiological data of help in distinguishing patients with bacterial coinfection. While the chest X-ray findings are nonspecific, they sometimes may provide an orientation as to the possible existence of bacterial coinfection. Specifically, if the interstitial lung condensations are accompanied by images of alveolar occupation (nodular or pseudonodular) or cavitations, the presence of bacterial coinfection should be suspected, particularly in the form of S. aureus. Although the presence of lobar or segmental infiltrates is suggestive of bacterial infection (despite their limited specificity), there have been descriptions of lobar condensations exclusively attributable to viral pneumonia due to influenza A/H1N1.101,106

Some biomarkers, especially procalcitonin, could be useful for deciding the start or early suspension of antibiotic treatment (ATB) in the case of negative bacterial cultures and antigenuria. Recently, a French multicenter study has suggested that procalcitonin <0.8μg/L offers a sensitivity of 91% and a specificity of 68%, with a negative predictive value of 91%, in reference to bacterial coinfection. Therefore, the absence of lobar condensations and a procalcitonin concentration of <0.8μg/L, together with the clinical course of the patient, could be used to decide the early discontinuation of ATB in the presence of negative microbiological findings.107

In adult patients, the empirical antibiotic treatment would be that recommended by the national guides of the SEIMC/SEMICYUC/SEPAR.108,109 In the case of patients with community-acquired pneumonia admitted to the ICU, the recommendation is a third generation cephalosporin (ceftriaxone or cefotaxime) plus a macrolide (some data suggest that the use of macrolides exerts an immune modulating effect, with an increase in the respiratory mucosal secretion of antiviral Ig A) or a fluoroquinolone. In the case of images suggestive of cavitation, or where gram staining shows a predominance of staphylococcus-type grampositive cocci, a switch to cloxacillin would be advisable. In Spain, empirical coverage for methicillin-resistant S. aureus (MRSA) would not be needed, except when the patient lives in a home for the elderly/sociosanitary center, or is known to have been previously colonized by MRSA. In such cases it would be advisable to add linezolid to the empirical treatment. In the case of patients with COPD, bronchiectasis, cystic fibrosis, or previous Pseudomonas aeruginosa colonization, we should administer an antipseudomonal betalactam (cefepime or piperacillin–tazobactam) instead of a third generation cephalosporin, associated to another ATB with action against P. aeruginosa, plus the macrolide.

The mortality rate among patients with influenza A/H1N1 infection who require admission to hospital is in the order of 6–7% vs 18–41% in those requiring admission to the ICU. Specifically, in the series of patients admitted to the ICU in Spain, the mean mortality rate among those patients requiring ventilation support was 30%.11,21,27,72

There is some controversy regarding the impact of bacterial coinfection upon the prognosis of patients with influenza A/H1N1 infection. The postmortem studies made in patients affected by other pandemics, as well as the recently published information referred to patients with influenza A/H1N1 infection,100 have shown that an important number of deaths attributed to the viral infection corresponded to patients who simultaneously presented bacterial infections. However, in epidemiological studies such as the Spanish multicenter survey that analyzed the impact of coinfection upon mortality in the ICU during the recent pandemic, multivariate analysis failed to identify the presence of coinfection as an independent factor correlated to mortality—though it was associated to a significant increase in the duration of stay in the ICU, and to increased resource consumption (days on mechanical ventilation and use of vasoactive drugs).75 Palacios et al.110 have published similar results. The authors found coinfection with S. pneumoniae to be associated to increased severity. This apparent discrepancy could be related to the limitations of postmortem studies in assessing the impact of a disease, since they only express the frequency of presentation of a given disorder in samples that are usually not representative of the population under study.

In patients with pneumonia due to influenza A/H1N1, and given the possibility of bacterial coinfection, the prescription of empirical antibiotic coverage is advised (associating a betalactam with a macrolide), administered as soon as possible. The results of the cultures and the clinical or laboratory test findings will serve to decide suspension or continuation of the antibiotics. As a preventive measure, antipneumococcal vaccination in the population at risk is advised.

On the basis of the existing information, there is not enough scientific evidence to recommend the use of corticosteroids in patients with viral pneumonia due to influenza A/H1N1, and we must wait for conclusive data from randomized, double-blind trials that avoid the mentioned confounding factors. However, the only such study in course (Low Dose Corticosteroids as Adjuvant Therapy for the Treatment of Severe H1N1 Flu) has been suspended because of low patient recruitment rates.

The role of noninvasive ventilation (NIV) in these patients has been the subject of controversy. On the basis of the evidence available at the time, the recommendation of the contingency plan of the SEMICYUC was to not use NIV in patients with acute respiratory failure secondary to ARDS due to influenza virus A/H1N1 infection, except in isolated cases. Approximately 30% of these patients were initially subjected to NIV, with a high failure rate (75%). The mortality rate among the patients in which NIV failed was 38%, and delays in intubation were associated to an increased mortality risk (OR=1.23; 95%CI 0.8–2.0).133 After carefully analyzing the results, it can be concluded that NIV possibly cannot be regarded as a technique of choice in these patients with ARDS, though it might be useful in centers that have great experience with the technique, and possibly using helmet-type interfaces134 which have been reported to offer good results, though in a very limited number of patients (n=5). In any case, we should follow the general recommendation to secure early intubation in these patients in response to the slightest evidence of NIV failure. This type of support would be viewed differently if the presentation of the disease were not in the form of ARDS.

There is general agreement on the use of “lung-protecting ventilatory strategies” when mechanical ventilation is invasive. The general principle of these strategies is to optimize the ventilatory parameters so as to avoid the cyclic collapse and opening of the sealed alveolar units as far as possible, and thus also prevent excessive stretching or over-distension of the healthy lung parenchyma. The mechanisms causing damage are shown in Table 5, and the basic principles of protective ventilation destined to minimize all these mechanisms are described in Table 6. Such strategies should be used early in all patients at risk of suffering acute lung injury, in order to avoid lung damage secondary to progression. The mechanical ventilation strategy applied in patients with ARDS in the influenza A/H1N1 pandemic was to use a low circulating volume associated to high positive end-expiratory pressure (PEEP) values, thereby “keeping the lung open”.135 In studies published in Canada and the United States,11,41 between 68 and 80% of these patients were ventilated with controlled modes, involving a circulating volume of approximately 6ml/kg, and a pause pressure target of under 35cmH2O—though some authors report that in a number of cases the circulating volume had to be increased to 8ml/kg, especially in patients with severe hypoxemia. Use was also made of high PEEP values of up to 30cmH2O, which were maintained for a long time, since it was seen that even in the respiratory disconnection phase the patients were very sensitive (in terms of hypoxemia) to even small reductions in PEEP. A special characteristic of this population was the high ventilatory demand, with great difficulties in keeping the patients adequately adapted to the respirator—requiring abnormally high doses of sedative and neuromuscular blockers for prolonged periods of time.

Pulmonary recruitment and ventilation in prone decubitus (which combines lung recruitment with improved elimination of the inflammatory lung fluids, avoiding damage secondary to spread) form part of the usual treatment of patients with refractory hypoxemia. Lung recruitment maneuvering, in lungs amenable to recruitment, produces the re-expansion of previously collapsed pulmonary zones via a brief and controlled increase in transpulmonary pressure. The problem is that such maneuvering produces a variable response dependent upon a series of factors such as the type of lesion, the evolutive phase of the lung damage, the severity of the lesion, the prior lung volume history, and the recruitment maneuver applied.136 Recently, systematic reviews have been made137,138 of studies that sum 1170 patients with ARDS, showing significantly improved oxygenation with the use of recruitment maneuvers, but with no evidence of any decrease in mortality (RR=0.73; 95%CI 0.46–1.7), and there moreover are not enough data to analyze their effect upon the duration of mechanical ventilation or hospital stay. Therefore, it can be affirmed that lung recruitment maneuvers, if performed adequately and with sufficient transpulmonary pressure, can improve PaO2 in severely hypoxemic patients, though it is not possible to recommend their generalized use in patients with ARDS. In the case of ARDS due to influenza A/H1N1 infection, there have been individual case descriptions documenting the success of lung recruitment in improving oxygenation, but it is not possible to draw firm conclusions regarding its benefit in this particular patient population. In view of the characteristics of the disorder, it can be affirmed that ARDS due to influenza A/H1N1 infection is associated to important alveolar flooding, with the consequent increase in lung weight—giving rise to models of easily recruitable distress which therefore might benefit from recruitment maneuvers and the use of high PEEP values.

Regarding ventilation in prone decubitus, it can be concluded that there is enough evidence of its efficacy in patients with severe ARDS in terms of both improved oxygenation and a significant decrease in mortality, without increasing the risk of serious complications.139 During the influenza A/H1N1 pandemic, many cases and series were published involving this technique, with good results.140 At present, the use of high-frequency ventilation requires further information in order to improve the existing level of evidence with a view to drawing reliable conclusions. Although a recent metaanalysis published by Sud et al.,141 involving 6 studies and including approximately 190 patients per management arm, has described a significant reduction in mortality after 30 days (RR=0.77; 95%CI 0.61–0.98), more evidence is needed, since this metaanalysis was based on studies in reference centers, not blinded studies, and with a small number of patients. A clinical trial with a larger sample size is currently pending publication and may yield new data with which to establish recommendations.

Another therapeutic measure to be considered in refractory hypoxemia, when pulmonary arterial hypertension is moreover detected, may be inhalatory nitric oxide (NO), which is associated to transient improvements in oxygenation. However, no investigation group has clearly shown NO to improve the patient prognosis.142

The management of ARDS in patients with influenza A/H1N1 infection must be based on lung-protecting ventilation strategies (tidal volume <10ml/kg and plateau pressure <35mmHg) and on the use of high PEEP values adjusted to the lung mechanics of the patient, combined with ventilation in prone decubitus, muscle relaxation and recruitment maneuvers. NIV cannot be regarded as an option of choice in patients with ARDS, though it could prove useful in centers with great experience and in cases of respiratory failure associated to COPD exacerbation or heart failure. Thus, the recommendations of the 2009 contingency plan of the SEMICYUC remain applicable.

A number of causes of cardiac involvement have been described in these patients:

• (a)

  Cardiac dysfunction secondary to ARDS: hypoxemia and pulmonary hypertension induce right ventricle dilatation, and in some cases dysfunction of both ventricles. Referred to by some authors as acute cor pulmonale,144,145 such dysfunction is observed in 25% of the patients and tends to be reversible as ARDS improves.

• (b)

  Myopericarditis: this disorder is the result of direct viral action upon the myocardium. The prevalence of myocarditis in influenza infections varies between 0 and 11%, depending on the definition used.146 In infection due to influenza A/H1N1, many cases of fulminant myopericarditis have been reported in young patients without risk factors, and who required inotropic drugs and vasopressors, and even ventricular assist systems. The clinical picture is no different from that of myocarditis of other causes. Following the viral episode, the patients develop clinical signs of heart failure, together with hypotension and tachycardia. In the more serious cases the condition can evolve towards refractory cardiogenic shock and death.147 The myocardial involvement tends to be reversible provided cardiocirculatory support measures are adopted, together with early antiviral treatment in the form of oseltamivir.

• (c)

  Exacerbation of pre-existing cardiovascular diseases: the influenza virus can cause alterations in endothelial function secondary to the systemic inflammatory process, inducing a procoagulant situation. The viral infection is also associated to atheroma plaque development and instability, which in turn can give rise to coronary ischemic episodes.148 A history of heart disease was present in 6–15% of the patients with influenza A/H1N1 infection during the pandemic of 2009.21 These patients may suffer destabilization as a result of the viral infection, with worsening of both systolic and diastolic ventricular function.

The diagnosis is based both on the clinical manifestations and on the complementary explorations and tests. The clinical condition is attributable to cardiac involvement, with a broad range of manifestations—though dyspnea is the most frequent symptom. Lower limb edemas or ascites are rare, and tend to manifest in patients with previous heart disease. The appearance or exacerbation of dyspnea, as well as the presence of signs of heart failure should cause us to suspect concomitant myocardial dysfunction.

Electrocardiogram: electrocardiographic alterations have been described in a large proportion of patients. The most common findings are T-wave inversion and ST-segment depression. These changes are usually transient and are not always correlated to elevations in cardiac enzymes and/or echocardiographic alterations. It is also possible to observe tachyarrhythmias (particularly supraventricular tachyarrhythmias), with a lesser frequency of bradyarrhythmias.149

Myocardial enzymes: troponin elevation is seen in up to 46% of the patients admitted to the ICU. However, not in all cases does this have an impact upon contractile function as determined by echocardiography. Pro-type B natriuretic factor is increased in the presence of both right and left ventricular dysfunction of any cause. Its determination is useful for stratifying the prognosis, evaluating treatment efficacy, and for identifying subclinical cardiac decompensation or the need for invasive hemodynamic monitorization.

Echocardiography is the technique of choice for the early diagnosis and management of patients with myocardial dysfunction. The most widely used parameters are the left ventricle ejection fraction (LVEF), the tricuspid annular plane systolic excursion (TAPSE) for assessing right ventricle function, and tissue Doppler ultrasound for the simultaneous evaluation of systolic and diastolic function. With tissue Doppler, 54% of the patients with troponin elevation or who require vasoactive drugs presented alterations in left ventricle function.150 Right-side dysfunction associated to ARDS is present in 8% of the cases.146 Pericardial effusion is not common, and when present it tends to be mild. Serial echocardiography in turn allows us to evaluate the persistence of alterations or the recovery of cardiac functions in a secure and noninvasive manner, and therefore to adjust drug treatment and circulatory and/or respiratory support, where considered appropriate.

Taking into account that there are no recommendations based on adequate scientific evidence, the evaluation and introduction of therapeutic measures can be made following the indications given in Fig. 5. Although the direct action of antiviral drugs upon the myocardium is not well known, the early start of antiviral treatment with oseltamivir is associated to an increased survival rate. It is essential for all patients with suspected influenza A/H1N1 infection to be treated immediately.

Figure 5.

Algorithm for the diagnosis and treatment of cardiovascular complications in patients with influenza A/H1N1 infection. IACB: intraaortic counterpulsation balloon; ECG: electrocardiogram; ECMO: extracorporeal membrane oxygenation; LVEF: left ventricle ejection fraction; Pro-BNP: type B natriuretic peptide; ACS: acute coronary syndrome; ARDS: acute respiratory distress syndrome; DICM: Department of Intensive Care Medicine; TAPSE: tricuspid annular plane systolic excursion; RV: right ventricle; LV: left ventricle.

(0.59MB).

It is advisable to initially evaluate cardiac function by echocardiography, and to repeat the exploration in the case of sudden or progressive clinical worsening. Left and right ventricle function should be assessed based on the LVEF and TAPSE, respectively, while systolic and diastolic function should be examined by tissue Doppler ultrasound. Depending on the clinical context of the patient, and after adequate hemodynamic monitorization, the most appropriate treatment should be started, based on inotropic and/or vasopressor drugs. Only in cases refractory to conventional treatment should the placement of an intraaortic counterpulsation balloon (IACB) be considered, and if this fails we should resort to a circulatory and/or respiratory assist device such as the extracorporeal membrane oxygenation (ECMO) technique. The use of IACB and of ECMO circulatory and/or respiratory assist devices complicates adequate hemodynamic monitorization, and it is not advisable to use monitoring systems based on analysis of the pulse wave profile, since the application of these circulatory assist devices would alter the measurements obtained. Most patients show recovery of cardiac and respiratory function with conventional treatment, without the need for assist devices.

Cardiovascular involvement in influenza A/H1N1 infection is frequent and secondary to the destabilization of pre-existing heart conditions including myocarditis, ischemic heart disease and right ventricle dysfunction. Early diagnosis and adequate monitorization allow us to start effective treatment and, in the more serious cases, assess the need for circulatory support systems.

To date, only three randomized studies have compared the use of ECMO vs conventional mechanical ventilation, and none of them included patients with influenza A/H1N1 infection (Table 7). The first two studies, published by Zapol et al.153 under the auspices of the United States National Institutes of Health (NIH) and by Morris et al.,154 revealed no reduction in hospital mortality with the use of ECMO. However, following the introduction of technological improvements in the pumps, oxygenators and cannulas of the ECMO circuit, the CESAR study,155 published in 2009, evidenced a decrease in mortality, with values at the limit of statistical significance. Nevertheless, on jointly assessing mortality and severe disability after 6 months, the difference was clearly favorable to ECMO (37% vs 53%; RR 0.69; 95%CI 0.05–0.97).

Regarding patients with influenza A/H1N1 infection, only more or less extensive case series have been published to date, reporting hospital mortality rates in patients subjected to ECMO of between 29 and 55%. The level of evidence supporting the utilization of ECMO therefore corresponds to class III, and the degree of recommendation is weak.

With the existing level of evidence, and considering the invasiveness of the technique and its potential complications, ECMO should be regarded as a rescue option in patients in which less invasive treatments have failed (e.g., optimization of mechanical ventilation and of alveolar recruitment, ventilation in prone decubitus, the use of nitric oxide, the administration of prostaglandins, or the use of high-frequency ventilation techniques where available).152 Its application must be restrictive, and ECMO therefore should only be indicated in the more seriously ill patients. The ECMO assessment and application guidelines are described in Table 8. Its use should be considered in those patients in which the estimated mortality is 50%, based on compliance with at least one of the criteria described in Table 8. In patients where the estimated mortality is 80%, immediate ECMO is indicated. Hemmila et al.,161 in a very large series, described greater mortality in those cases where before the application of ECMO the patients suffered severe acidosis with pH<7.10 (OR=8.4), or had required mechanical ventilation during >8 days (OR=5.5). In addition to the impossibility of securing systemic anticoagulation, the absolute contraindications of ECMO are those derived from terminal situations or previously greatly deteriorated patient conditions, as well as established multiorgan failure (MOF). In the case of venous–arterial ECMO, severe aortic insufficiency also should be regarded as an absolute contraindication. The relative contraindications of the technique in turn are derived from the poorer results obtained in elderly patients and obese individuals, and when ECMO use is delayed beyond 7 days of mechanical ventilation with high FiO2 and PEEP requirements.

In view of its complexity, with important instrumental, technical, training and qualified healthcare personnel requirements, ECMO preferentially should be available in third level hospital centers. Simultaneously, and in order to comply with the ethical norms referred to the equity of access to medical care, it is advisable to incorporate ECMO to reference hospitals, with a multidiscipline team composition in which an intensivist must always be present. The function of this team is to establish consensus-based protocols and to jointly evaluate with the physician caring for the patient the advisability of transferring the patient to the reference hospital for ECMO implantation. This type of centralized structure is effective in attending patients admitted to hospitals where ECMO is not available, and has been adopted with good healthcare, training and resource optimization results in countries such as Australia–New Zealand156 and Italy.157 It is also recommended by the Extracorporeal Life Support Organization160 in the United States. Based on these premises, Fig. 6 shows the management of patients with ARDS secondary to influenza A/H1N1 infection, according to the type of hospital in which they are admitted.

Figure 6.

Algorithm for the management of patients with refractory acute respiratory distress syndrome secondary to influenza A/H1N1 infection, according to the level of hospital care. ECMO: extracorporeal membrane oxygenation; PEEP: positive end-expiratory pressure; ARDS: acute respiratory distress syndrome.

(0.42MB).

ECMO is a rescue technique in cases of influenza A/H1N1 infection with refractory ARDS. The supporting scientific evidence is weak, and ECMO is not a first choice technique. It is used when other less invasive options have failed. It should be applied in patients with a poor prognosis (estimated mortality >80%). The most effective and efficient model comprises centralization in reference hospitals and the creation of multidiscipline ECMO teams capable of implanting the technique in other hospital centers and of transferring patients. The clinical results obtained in large series reflect in-hospital survival rates of approximately 50–60%.

Vaccination is one of the most effective measures for the prevention and control of certain diseases. In this sense, the best way to prevent influenza A/H1N1 infection is undoubtedly annual influenza vaccination comprising the updated prevalent circulating strains of the virus in each given season.

The pandemic spread of the new variant of influenza A/H1N1 during the year 2009, influenza A/California/7/2009 (H1N1), made it necessary to quickly prepare a monovalent vaccine including only the attenuated new strain, followed by inclusion of the latter in the formulation of the annual influenza vaccines (trivalent vaccines), starting with the season 2010–2011. In addition, the epidemiological data obtained since then have led to revision and updates of the vaccination recommendations on the part of most of the implicated healthcare organisms.

In Europe, and in relation to seasonal influenza, use is only made of inactivated vaccines often formulated with adjuvants. Most are derived from embryonated chicken eggs via fractions of the virus obtained after treatment with detergents (split vaccines), or antigens or purified subunits. For patients allergic to egg proteins we also have vaccines obtained from viruses replicated in cell cultures. In the United States, the available inactivated vaccines do not use adjuvants. An attenuated live vaccine is also available, administered via the intranasal route (FluMist®). Recently, inactivated vaccines without adjuvants have been marketed, designed for intradermal administration in adults between 18 and 64 years of age, and which induce an equivalent immune response with a lesser amount of antigen and a smaller administered volume (0.1ml) than the conventional vaccines. In the United States, an inactivated vaccine has also been approved with a four-fold higher antigen concentration than the conventional vaccine (Fluzone—“High-Dose”®), for the immunization of individuals over 65 years of age. In the second half of the year 2009, the European Medicines Agency (EMA, previously EMEA) authorized three monovalent inactivated vaccines based on the pandemic influenza A/California/7/2009 (H1N1) strain: two with an oily adjuvant (scualene) based on recombinant virus (Pandemmrix® from GSK and Focetria® from Novartis), and a third vaccine (Celvapan® from Baxter) without adjuvant and which uses the complete virus. Posteriorly, and in the context of an emergency procedure, two other inactivated pandemic vaccines with oily adjuvants were authorized: Arepanrix®, produced by GSK, and Humenza®, from Sanofi-Pasteur. The pandemic vaccines were administered in a single dose during the 2009 season, together with the seasonal influenza vaccine. Starting in 2010 and 2011, the pandemic strain has been included in the trivalent seasonal influenza vaccines, together with antigens derived from the A/Perth/16/2009 (H3N2) and B/Brisbane/60/2008 strains in the northern hemisphere for the 2011–2012 season.162

The inactivated vaccine is administered via the deep intramuscular route, in the deltoid muscle. In nursing infants and small children, injection preferably should be made in the lateral zone of the thigh. In the northern hemisphere, the first cases of seasonal influenza can appear in November; as a result, the vaccination campaigns are generally activated in the month of October. The inactivated vaccine is safe in pregnancy, and its administration (preferably without adjuvant) is recommended in any stage of pregnancy in women with high-risk medical conditions, and from gestational week 14 in the rest of women. The attenuated live vaccine is administered via the nasal route and can cause mild respiratory symptoms. It can be used in healthy patients between 2 and 49 years of age. The vaccine is not recommended in patients with comorbidities, and is contraindicated both in severely immune depressed patients and in people living with them. In general, simultaneous administration of the seasonal influenza vaccine with other vaccines (whether live or inactivated) is not advised, with the exception of the antipneumococcal vaccine, which when indicated can be coadministered in the contralateral deltoid muscle.

The response to vaccination is detected by measuring the serum titers of neutralizing antibodies and hemagglutination inhibiting antibodies in the vaccinated patient. Hemagglutination inhibiting antibody titers of over 1:32–1:40 are considered adequate. With the trivalent vaccines, protective antibodies are present from two weeks after vaccination, with reductions to below protective titers after 6–8 months—this in turn justifying annual revaccinations.163 The studies made with the pandemic A/H1N1 vaccine of 2009 showed children under 9 years of age to present lower response rates than older children and young adults. As a result, the United States Advisory Committee on Immunization Practices (ACIP) recommends that children between 6 months and 8 years of age vaccinated for the first time, and those who during the previous season had received a single dose, should be administered a second dose four weeks after the first. Immune depressed individuals show lower response rates than the general population, though immunization in these subjects does not improve with revaccination; as a result, neither the administration of two doses nor an increase in dosage is indicated in such cases.

The advisory committees of the United States and Canada (ACIP/NACI) currently recommend universal influenza vaccination from 6 months of age, except in the presence of contraindications. In most countries vaccination is indicated in risk groups, following the recommendations of the WHO,162 which prioritizes the vaccination of elderly people, institutionalized persons, those with chronic diseases, pregnant women, healthcare workers, children between 6 months and two years of age, and people with functions essential to society. In any case, annual revaccination is advised in order to ensure optimum protection, despite the fact that the vaccine may contain the same strains as in the previous season, since host immunity decreases in the months after vaccination or in natural infection with the influenza virus. This decrease in turn is faster in elderly people and in patients with comorbidities. The recommendations of the Interterritorial Council of the Spanish Ministry of Health164 referred to influenza vaccination are shown in Table 9.

In the systematic review of the clinical trials with vaccines against seasonal influenza conducted in healthy adults and children, the efficacy in preventing influenza virus infection confirmed by laboratory testing was 70–90%, though the percentage may vary depending on the discrepancies between the vaccinal and circulating viral strains.165 In the general population, the clinical efficacy of the vaccine can be estimated by statistical case–control analyses of the incidence of influenza-like syndromes confirmed by laboratory techniques and which have required medical care. By means of the sentinel physician networks, the Influenza Monitoring Vaccine Effectiveness in Europe project has estimated the adjusted effectiveness of the trivalent inactivated vaccine for the 2010–2011 season. The recorded effectiveness was moderate: 42.3% (95%CI 7.3–69.0%), which is lower than the values obtained in the randomized clinical trials made in selected healthy populations.166 With due consideration of the limitations inherent to large observational studies, the cohorts studied in the United States in the 1990s showed influenza vaccination of the elderly population over 65 years of age to have an impact not only on respiratory disease—with a 29–32% reduction in hospital admissions due to pneumonia or flu167—but also on cardiac and cerebrovascular mortality,168 with a 48–50% reduction in global mortality of all causes.

No vaccine is absolutely safe. Moreover, paradoxically, when thanks to the efficacy of a vaccine concern about the disease it helps control decreases, fear of the risks associated to vaccine use is seen to increase. In this sense, continuous monitorization of vaccine safety is needed in order to ensure that the population continues to trust the vaccination programs. In the case of the influenza A/H1N1 vaccine, doubts about its efficacy and possible adverse effects have been one of the main causes limiting its use. Even among healthcare workers, which theoretically represent one of the best informed risk groups, low adherence to the vaccination campaigns has been documented.169 Furthermore, in centers with well informed and aware healthcare personnel, the influenza A/H1N1 vaccination rate was lower than for seasonal influenza (75.2% vs 93.8%).170 The vigilance systems that monitor the safety of the influenza A/H1N1 vaccines (both with and without adjuvants) generally report adverse events (AEs) rates similar to those recorded for seasonal influenza. The published AE rates vary greatly (0.2–41.7% of all vaccinated individuals), depending particularly on the study population and on the vigilance system involved.171,172 The incidence density varies between 9 and 31.8 AEs for every 100,000 administered doses. Problems in this sense are more frequent in the first 24h after vaccination in the younger population, and there are no gender differences.172,173 The most common AEs are considered mild (fever [15%], reddening-induration-local pain [38%], muscle pain [5–7%], rhinorrhea [10–15%], headache [12–27%], hysteria reactions), with incidence densities between 1.9 and 0.1 episodes/100,000 vaccinations,174 and which in some series represented over 90% of the total AEs.174 In addition, there have been reports of different types of allergic reactions, urticaria (10%), angioedema (0.5%), Schönlein-Henoch purpura (0.9%), anaphylaxis (0.6%), etc. On a point basis there have been reports of hematological AEs in the form of aplastic anemia or severe thrombocytopenia, and even episodes of sudden death.175 However, the AEs that have generated most concern are of a neurological nature, with isolated cases of encephalomyelitis,173,174 and particularly Guillain–Barré syndrome (GBS) and narcolepsy.

In the case of GBS, a number of studies have examined its possible association to the monovalent vaccine, though no consistent relationship has been demonstrated.173–178 Either no increase in the incidence of GBS after vaccination has been observed (31 expected cases; 7 observed cases), or the risk seems smaller with the influenza A/H1N1 vaccine than with the seasonal influenza vaccine (adjusted OR=0.9 vs 1.3). Other findings have been that both influenza infection and other respiratory infections and the seasonal vaccine constitute risk factors associated to the appearance of GBS, acting as confounding elements on analyzing the effects of the influenza A/H1N1 vaccine (non-adjusted OR=2.8; 95%CI 1.3–6; adjusted OR=1; 95%CI 0.3–2.7). Likewise, it has not been possible to identify a clear biological basis confirming the possible epidemiological link. On the other hand, following the 2009–2010 vaccination campaign, an abnormally large number of narcolepsy episodes were recorded in patients between 4 and 19 years of age in Finland and Sweden, and to a lesser degree in France.179 The estimated risk was 1 case/12,000 vaccinations, with a mean time to appearance of 52 days. This phenomenon appears to be associated with the monovalent vaccines containing adjuvant AS03. As a result, the EMEA (21 July 2011), while considering the overall risk-benefit ratio of this group of vaccines to be positive, recommended avoiding the Pandemmrix® vaccine in patients under 20 years of age, except when the trivalent seasonal influenza vaccine is not available.180 At present, studies are nearing completion that are hoped to be able to clarify this problem. Lastly, it must be remembered that pregnant women are at a special risk of developing complications associated to influenza, and although vaccination is advised in these cases, the observed vaccination rate is notoriously low (12–24%).181 However, although the available information is limited, the influenza A/H1N1 vaccine does not appear to increase the risk of maternal-fetal complications during pregnancy. The most common AE is spontaneous miscarriage (41.2%), followed by fetal death (6.5%).182 The most frequent AEs unrelated to pregnancy are non-anaphylactic allergic reactions, other local reactions, and the appearance of general symptoms. However, the most relevant finding is that there have been no unusual complication profiles in the pregnant women or in the fetuses. The rates corresponding to spontaneous miscarriage, fetal death, congenital anomalies and other complications appear to be no different from those described in the general population.183,184

Chemoprophylaxis must always be regarded as a complement to vaccination. The use of adamantanes (amantadine and rimantadine) is not indicated, due to the resistance rates observed in the influenza A/H1N1 isolates. With the neuraminidase inhibitors (zanamivir or oseltamivir), the reported efficacy rates are 70–90%. Such therapy is indicated in patients at a high risk of developing complications due to influenza and who cannot receive the vaccine.185 In addition, it is particularly useful in high risk patients and in healthcare personnel who have suffered exposure during epidemic outbreaks (duration of 10 days in the case of home exposure, and 7 days in others). A minimum of two weeks of chemoprophylaxis is advised for the control of outbreaks, even in vaccinated individuals, maintaining treatment up to one week after identification of the last case. Chemoprophylaxis should not be administered during 48h before or two weeks after administration of the live-attenuated intranasal vaccine. In contrast, chemoprophylaxis has no effect upon the inactivated vaccine.

Influenza vaccination is advised of all patients at risk, though it may be necessary to extend this indication to all patients over 6 months of age, except in the presence of contraindications. Children should receive two doses spaced one month apart. Immune depressed individuals and the population at risk should receive one dose, with annual revaccination. The frequency of adverse events following administration of the influenza A/H1N1 vaccine is similar to that seen with seasonal influenza vaccination. Chemoprophylaxis must always be regarded as a complement to vaccination, and is indicated in individuals at high risk of developing complications, and in healthcare workers who have suffered exposure.