Professional Documents
Culture Documents
Non-Invasive Ventilation in Community-Acquired Pneumonia and Severe Acute Respiratory Failure
Non-Invasive Ventilation in Community-Acquired Pneumonia and Severe Acute Respiratory Failure
Non-Invasive Ventilation in Community-Acquired Pneumonia and Severe Acute Respiratory Failure
net/publication/221817551
CITATIONS READS
146 1,138
8 authors, including:
Antonio Torres
IRB Barcelona Institute for Research in Biomedicine
1,078 PUBLICATIONS 42,068 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Andres Carrillo on 30 December 2013.
Andres Carrillo
Gumersindo Gonzalez-Diaz
Non-invasive ventilation in community-acquired
Miquel Ferrer pneumonia and severe acute respiratory failure
Maria Elena Martinez-Quintana
Antonia Lopez-Martinez
Noemi Llamas
Maravillas Alcazar
Antoni Torres
Received: 22 July 2011 Abstract Purpose: The use of independently predicted NIV failure.
Accepted: 6 January 2012 non-invasive ventilation (NIV) in Likewise, maximum SOFA, NIV
Published online: 9 February 2012 severe acute respiratory failure (ARF) failure and older age independently
! Copyright jointly held by Springer and due to community-acquired pneumo- predicted hospital mortality. Among
ESICM 2012 intubated patients with ‘‘de novo’’
nia (CAP) is controversial, and the
risk factors for NIV failure in these ARF, NIV duration was shorter in
patients are not well known. We hospital survivors than non-survivors
assessed the characteristics and pre- (32 ± 24 versus 78 ± 65 h,
dictors of outcome of patients with p = 0.014). In this group, longer
CAP and severe ARF treated with duration of NIV before intubation
A. Carrillo ! G. Gonzalez-Diaz ! NIV. Methods: We prospectively was associated with decreased hospi-
M. E. Martinez-Quintana ! assessed 184 consecutive patients; tal survival (adjusted odds ratio
A. Lopez-Martinez ! N. Llamas ! 102 had ‘‘de novo’’ ARF, and 82 0.978, 95% confidence interval
M. Alcazar 0.962–0.995, p = 0.012). This asso-
ICU, Hospital JM Morales Meseguer,
previous cardiac or respiratory dis-
Murcia, Spain ease. We defined successful NIV as ciation was not observed in patients
avoidance of intubation and intensive with previous cardiac or respiratory
M. Ferrer ()) ! A. Torres care unit (ICU) survival at least 24 h disease. Conclusions: Successful
UVIIR, Servei de Pneumologia, Institut del in the ward. We assessed predictors NIV was strongly associated with
Tòrax, Hospital Clinic, IDIBAPS, of NIV failure and hospital mortality better survival. If predictors for NIV
Universitat de Barcelona, Villarroel 170, in multivariate analyses. Results: failure are present, avoiding delayed
08036 Barcelona, Spain intubation of patients with ‘‘de novo’’
e-mail: miferrer@clinic.ub.es
Patients with ‘‘de novo’’ ARF failed
Tel.: ?34-93-2275549 NIV more frequently than patients ARF would potentially minimise
Fax: ?34-93-2275549 with previous cardiac or respiratory mortality.
URL: http://www.idibapsrespiratoryresearch. disease (47, 46% versus 21, 26%,
org p = 0.007). Worsening radiologic Keywords Non-invasive ventilation !
infiltrate 24 h after admission, maxi- BiPAP ! Acute respiratory failure !
M. Ferrer ! A. Torres mum Sepsis-Related Organ Failure Severe community-acquired
Centro de Investigación Biomedica En Red- pneumonia
Enfermedades Respiratorias (CibeRes,
Assessment (SOFA) score and, after
CB06/06/0028)-Instituto de Salud Carlos 1 h of NIV, higher heart rate and
III, Madrid, Spain lower PaO2/FiO2 and bicarbonate
aetiologies [7–9]. The most important rationale for using chronic respiratory diseases, or in the presence of
NIV is to overcome an episode of severe acute respiratory wheezing received nebulised bronchodilators and sys-
failure (ARF) without the need for endotracheal intuba- temic steroids. Patients with hypotension received
tion. However, few studies have assessed specifically the initially intravenous fluid therapy with crystalloids; if
usefulness of NIV in patients with pneumonia [10], and it haemodynamic instability persisted, norepinephrin was
is even considered controversial due to a major variability initiated. Patients who responded to vasoactive drugs
in failure rates [10–13], which are generally higher than remained on NIV unless they required intubation for other
those observed in chronic obstructive pulmonary disease reasons.
(COPD) [14] or cardiogenic pulmonary oedema [15].
Moreover, predictive factors for NIV failure have been
studied in different aetiologies [13, 14, 16, 17]. A recent Non-invasive ventilation protocol
study found that lack of improvement in arterial oxy-
genation was the main predictor of NIV failure in patients The indication for NIV was done by the attending phy-
with CAP [18]. However, this study assessed a limited sicians. Patients were continuously monitored with
number of patients. Therefore, studies assessing factors electrocardiogram, pulse oximetry, invasive and/or non-
predicting outcome for NIV in larger populations of CAP invasive blood pressure, and respiratory rate. The criteria
with severe ARF are still needed. used for implementing NIV were: (1) moderate to severe
In spite of the controversy surrounding the use of NIV dyspnoea accompanied by respiratory rate C30 breaths/
out of COPD exacerbations or cardiogenic pulmonary min or signs of increased work of breathing such as active
oedema, several observational reports illustrate that hyp- contraction of the accessory respiratory muscles, and (2)
oxaemic ARF is among the most frequent indications for arterial oxygen tension to inspired oxygen fraction (PaO2/
NIV in real practice [19, 20]. The ICU of our institution FiO2) ratio \250. Patients in need for emergency intu-
has been using NIV for the management of patients with bation at admission (respiratory or cardiac arrest,
severe ARF secondary to pneumonia who do not need respiratory pauses with loss of alertness or gasping for air,
emergency intubation for more than 15 years. major agitation inadequately controlled by sedation, signs
Since the available evidence on the efficacy of NIV in of exhaustion, massive aspiration, inability to manage
pneumonia is not compelling, appropriate patient selec- respiratory secretions appropriately, and haemodynamic
tion should be a key point to improve the success of this instability without response to fluids and vasoactive
technique. To determine the clinical results and evolution agents) [13] were not treated with NIV.
of patients with CAP and severe ARF treated with NIV at Non-invasive ventilation was applied with specific
our institution, we assessed the characteristics and out- NIV ventilators (BiPAP ST-D and VISION Ventilator;
comes of these patients and determined factors that Respironics, Inc., Murrysville, PA, USA). As previously
predict failure of the technique and mortality. described [13, 21], patients were ventilated using the bi-
level positive airway pressure mode (BiPAP). Face mask
was used as first choice, but nasal mask was optionally
used if patients did not tolerate face mask. The initial
Methods inspiratory positive airway pressure (IPAP) was set at
12 cmH2O, and levels were raised by 2–3 cmH2O as
Patients tolerated but did not exceed 25 cmH2O. The initial
expiratory positive airway pressure (EPAP) was set at
We prospectively followed all consecutive patients with 5 cmH2O, and the levels were raised by 1–2 cmH2O if
severe ARF due to CAP receiving NIV treatment in an needed to improve hypoxaemia. The FiO2 was set to
18-bed ICU of Hospital Morales Meseguer, a general achieve SpO2 [92% or PaO2 [65 mmHg. Arterial blood
university hospital in Murcia, Spain, from January 1997 gas samples were obtained from each patient before the
to December 2008. The study was approved by the Ethics connection to the ventilator and after 1 h of NIV. Sub-
Committee of the institution. sequently, samples were obtained as clinically indicated.
Pneumonia was defined as a new pulmonary infiltrate
on the admission chest radiograph and symptoms and
signs of lower respiratory tract infection. Exclusion cri- Effectiveness of the technique
teria were any degree of immunosuppression and other
criteria previously published [21]. All patients received Non-invasive ventilation was considered successful if a
timely empiric antimicrobial therapy according to the patient avoided endotracheal intubation, was discharged
local protocols. from the ICU and remained alive and conscious for at
All patients received stress ulcer and deep venous least 24 h after being transferred to the ward. Failure
thrombosis prophylaxis. Patients with COPD, other of NIV therapy occurred when a patient experienced
460
Table 1 Clinical and ventilatory characteristics of patients with ‘‘de novo’’ acute respiratory failure and with previous cardiac or
respiratory disease
drugs at onset of NIV, severity scores and acidosis, had Analyses of survival (Table 4)
less extensive radiologic findings, less acute respiratory
distress syndrome (ARDS) criteria, and lower respiratory Hospital survivors from both groups were younger, less
and heart rate and better oxygenation after 1 h of NIV. In severe, assessed by severity scores and acidosis, had
patients with ‘‘de novo’’ ARF, successful NIV was also failed NIV less frequently, and had lower respiratory rate
associated with lower respiratory rate and better oxy- after 1 h of NIV. In patients with ‘‘de novo’’ ARF, sur-
genation at onset of NIV. In patients with previous vival was also associated with less worsening of
cardiac or respiratory disease, successful NIV was also pulmonary infiltrate 24 h after the onset of NIV and better
associated with lower heart rate at onset of NIV and oxygenation. In patients with previous cardiac or respi-
higher arterial pH and better level of consciousness after ratory disease, survival was also associated with bilateral
1 h of NIV. Age, sex and PaCO2 were not related with radiologic involvement and ARDS criteria, and need for
NIV outcome in both groups. vasoactive drugs at onset of NIV. As for NIV outcome,
The ICU stay was shorter, and mortality lower in sex and PaCO2 were not related with survival in both
patients with successful treatment with NIV. groups.
Since the variables significantly associated with NIV In the overall population, the multivariate analysis
outcome were essentially the same for patients with identified maximum SOFA score during the ICU stay,
‘‘de novo’’ ARF and previous cardiac or respiratory dis- NIV failure and older age as independent predictors of
ease (Table 2), the multivariate analysis of predictors of hospital mortality (Table 5). Maximum SOFA score and
NIV outcome was performed in the overall population. NIV failure were the most accurate variables in predicting
The variables independently associated with NIV failure NIV failure, as shown in this table.
were worsening of radiological infiltrate 24 h after
admission, maximum SOFA score during NIV, and higher
heart rate, lower PaO2/FiO2 ratio and lower bicarbonate Delayed intubation and hospital mortality
after 1 h of NIV (Table 3). Except for heart rate, these
variables were highly accurate in predicting NIV failure, Among intubated patients with ‘‘de novo’’ ARF, the
as shown in this table. duration of NIV in hospital survivors was shorter than that
462
Table 2 Variables associated with successful and unsuccessful treatment with non-invasive ventilation in patients with ‘‘de novo’’ ARF
and previous cardiac or respiratory disease
Table 3 Multivariate analysis of variables independently associated with non-invasive ventilation failure in the overall population
Maximum SOFA during NIV 1.442 1.187–1.753 \0.001 0.86 C7 81 80 4.08 0.24
Worsening X-ray infiltrate 24 h 84.23 16.74–423.8 \0.001 – – 77 86 5.58 0.27
after onset of NIV
Heart rate 1 h after NIV onset, min-1 1.064 1.029–1.100 \0.002 0.68 C104 63 67 1.93 0.55
PaO2/FiO2 ratio 1 h after 0.980 0.965–0.996 0.012 0.78 \144 53 91 5.58 0.52
NIV onset, mmHg
HCO3 1 h after NIV onset, mEq/L 0.802 0.711–0.905 \0.001 0.77 \23 67 68 2.72 0.48
Adj. OR adjusted odds ratio, CI confidence interval, SOFA Sepsis-Related Organ Failure Assessment, AUC area under the ROC curve
of non-survivors (32 ± 24 versus 78 ± 65 h, p = 0.014, that best predicted mortality was C53 h, with sensitivity
Fig. 1 left panel), with an AUC of 0.70 for predicting of 69% and specificity of 83%. The SAPS-II, CURB65
hospital mortality. The duration of NIV before intubation and SOFA scores at admission did not correlate with the
463
Table 4 Variables associated with hospital mortality in patients with ‘‘de novo’’ ARF and previous cardiac or respiratory disease
Table 5 Multivariate analysis of variables independently associated with hospital mortality in the overall population
Maximum SOFA during ICU stay 1.342 1.158–1.556 \0.001 0.86 C12 68 95 13.3 0.34
NIV failure 6.78 1.65–27.95 0.008 – – 79 77 3.48 0.27
Older age (years) 1.118 1.056–1.185 \0.001 0.68 C72 72 57 1.68 0.49
Adj. OR adjusted odds ratio, CI confidence interval, SOFA Sepsis-Related Organ Failure Assessment, AUC area under the ROC curve,
NIV non-invasive ventilation
duration of NIV in these patients, indicating that patients By contrast, no relationship was found between
with longer time on NIV before intubation were not more duration of NIV before intubation and mortality in
severely ill at admission. Moreover, the need for vaso- patients with previous cardiac or respiratory disease
active drugs at onset of NIV was associated with shorter (Fig. 1, right panel).
duration of NIV before intubation (31 ± 40 versus
65 ± 52 h, respectively, p = 0.030), indicating that
patients treated with vasoactive drugs who need intuba-
tion fail earlier than those without these drugs. After Discussion
adjusting for these variables, increased duration of
NIV before intubation in hours was significantly associ- Patients with CAP and previous cardiac or respiratory
ated with decreased hospital survival (adjusted odds disease responded better to NIV than those with
ratio 0.978, 95% confidence interval 0.962–0.995, ‘‘de novo’’ ARF, as expected from previous studies [10,
p = 0.012). 19]. Successful NIV treatment strongly predicted
464
"De novo" acute Previous cardiac or improvement of blood gases after ventilatory support
250 respiratory failure 125 respiratory disease were related with NIV failure on univariate analysis. This
study, however, was restricted to patients with ‘‘de novo’’
p=0.014 ARF, and did not consider variables other than general
p=0.68
200
severity scores and ventilatory and respiratory parameters.
100
Studies on NIV often include pneumonia as a cause of
‘‘de novo’’ or hypoxaemic ARF only. However, the great
proportion of previous cardiac or respiratory disease in
150 75 our patients is not surprising, since large published series
Hours
with ‘‘de novo’’ ARF, we strongly advise careful man- randomised controlled trial. In the absence of a control
agement in order to avoid potential harm related to group, definitive conclusions cannot be drawn. Second,
excessive delay in intubation when the response to NIV is the study was conducted in a centre with major experi-
not positive, particularly when they require vasoactive ence in the use of NIV, and therefore these results cannot
drugs, a condition particularly associated with NIV failure be extrapolated to less trained and equipped hospitals.
in patients with acute lung injury [32], as well as in those In conclusion, successful treatment with NIV, which is
patients who do not achieve those PaO2/FiO2 cut-offs strongly related with less organ system failure of patients
after 1 h of NIV treatment, as previously described [17, and a good initial response to the antimicrobial treatment
33]. By contrast, this recommendation cannot apply for and NIV support, was strongly associated with better
patients with previous cardiac or respiratory disease. survival. In presence of predictors for NIV failure,
In the overall population, older age, NIV failure and avoiding delayed intubation of patients with ‘‘de novo’’
higher organ system failure were independently associ- ARF is strongly advised in order to potentially minimise
ated with hospital mortality. Thus, further improvement mortality.
of survival could be potentially achieved by combining
adequate support against organ system dysfunction and Acknowledgment Funded by CibeRes (CB06/06/0028)-ISCiii,
optimal NIV management of these patients. 2009 SGR 911, IDIBAPS.
The present study has several limitations. First, the
effectiveness of any treatment must be determined by a
References
1. Mandell LA, Wunderink RG, Anzueto 10. Confalonieri M, Potena A, Carbone G, 16. Rodriguez ML, Carrillo AA, Melgarejo
A, Bartlett JG, Campbell GD, Dean NC, Della Porta R, Tolley E, Meduri G MA, Renedo VA, Parraga RM, Jara PP,
Dowell SF, File TM Jr, Musher DM, (1999) Acute respiratory failure in Millan MJ, Gonzalez DG (2005)
Niederman MS, Torres A, Whitney CG patients with severe community- Predictive factors related to success of
(2007) Infectious Diseases Society of acquired pneumonia. A prospective non invasive ventilation and mortality
America/American Thoracic Society randomized evaluation of noninvasive in the treatment of acute cardiogenic
consensus guidelines on the ventilation. Am J Respir Crit Care Med pulmonary edema. Med Clin (Barc)
management of community-acquired 160:1585–1591 124:126–131
pneumonia in adults. Clin Infect Dis 11. Jolliet P, Abajo B, Pasquina P, 17. Antonelli M, Conti G, Moro ML,
44(Suppl 2):S27–S72 Chevrolet JC (2001) Non-invasive Esquinas A, Gonzalez-Diaz G,
2. Restrepo MI, Anzueto A (2009) Severe pressure support ventilation in severe Confalonieri M, Pelaia P, Principi T,
community-acquired pneumonia. Infect community-acquired pneumonia. Gregoretti C, Beltrame F, Pennisi MA,
Dis Clin North Am 23:503–520 Intensive Care Med 27:812–821 Arcangeli A, Proietti R, Passariello M,
3. Oosterheert JJ, Bonten MJ, Hak E, 12. Domenighetti G, Gayer R, Gentilini R Meduri GU (2001) Predictors of failure
Schneider MM, Hoepelman AI (2003) (2002) Noninvasive pressure support of noninvasive positive pressure
Severe community-acquired ventilation in non-COPD patients with ventilation in patients with acute
pneumonia: what’s in a name? Curr acute cardiogenic pulmonary edema hypoxemic respiratory failure: a multi-
Opin Infect Dis 16:153–159 and severe community-acquired center study. Intensive Care Med
4. Pierson DJ (2002) Indications for pneumonia: acute effects and outcome. 27:1718–1728
mechanical ventilation in adults with Intensive Care Med 28:1226–1232 18. Carron M, Freo U, Zorzi M, Ori C
acute respiratory failure. Respir Care 13. Ferrer M, Esquinas A, Leon M, (2010) Predictors of failure of
47:249–262 Gonzalez G, Alarcon A, Torres A noninvasive ventilation in patients with
5. Chastre J, Fagon JY (2002) Ventilator- (2003) Noninvasive ventilation in severe community-acquired pneumonia.
associated pneumonia. Am J Respir Crit severe hypoxemic respiratory failure: a J Crit Care 25:514–540
Care Med 165:867–903 randomized clinical trial. Am J Respir 19. Demoule A, Girou E, Richard JC, Taille
6. Pinhu L, Whitehead T, Evans T, Crit Care Med 168:1438–1444 S, Brochard L (2006) Benefits and risks
Griffiths M (2003) Ventilator- 14. Lightowler JV, Wedzicha JA, Elliott of success or failure of noninvasive
associated lung injury. Lancet MW, Ram FS (2003) Non-invasive ventilation. Intensive Care Med
361:332–340 positive pressure ventilation to treat 32:1756–1765
7. Mehta S, Hill NS (2001) Noninvasive respiratory failure resulting from 20. Schettino G, Altobelli N, Kacmarek
ventilation (State of the Art). Am J exacerbations of chronic obstructive RM (2008) Noninvasive positive-
Respir Crit Care Med 163:540–577 pulmonary disease: cochrane systematic pressure ventilation in acute respiratory
8. (2001) International consensus review and meta-analysis. BMJ failure outside clinical trials: experience
conferences in intensive care medicine: 326:185–189 at the Massachusetts General Hospital.
noninvasive positive pressure 15. Weng CL, Zhao YT, Liu QH, Fu CJ, Crit Care Med 36:441–447
ventilation in acute respiratory failure. Sun F, Ma YL, Chen YW, He QY
Am J Respir Crit Care Med (2010) Meta-analysis: noninvasive
163:283–291 ventilation in acute cardiogenic
9. Nava S, Hill N (2009) Non-invasive pulmonary edema. Ann Intern Med
ventilation in acute respiratory failure. 152:590–600
Lancet 374:250–259
466
21. Gonzalez Diaz G, Carrillo Alcaraz A, 27. Fischer JE, Bachmann LM, Jaeschke R 33. Antonelli M, Conti G, Esquinas A,
Pardo Talavera JC, Jara Perez P, (2003) A readers’ guide to the Montini L, Maggiore SM, Bello G,
Esquinas Rodriguez A, Garcia Cordoba interpretation of diagnostic test Rocco M, Maviglia R, Pennisi MA,
F, Hill NS (2005) Noninvasive positive- properties: clinical example of sepsis. Gonzalez-Diaz G, Meduri GU (2007) A
pressure ventilation to treat hypercapnic Intensive Care Med 29:1043–1051 multiple-center survey on the use in
coma secondary to respiratory failure. 28. Lim WS, Baudouin SV, George RC, clinical practice of noninvasive
Chest 127:952–960 Hill AT, Jamieson C, Le JI, Macfarlane ventilation as a first-line intervention
22. Le Gall JR, Lemeshow S, Saulnier F JT, Read RC, Roberts HJ, Levy ML, for acute respiratory distress syndrome.
(1993) A new simplified acute Wani M, Woodhead MA (2009) BTS Crit Care Med 35:18–25
physiology score (SAPS II) based on a guidelines for the management of 34. Liapikou A, Ferrer M, Polverino E,
European/North American multicenter community acquired pneumonia in Balasso V, Esperatti M, Piner R, Mensa
study. JAMA 270:2957–2963 adults: update 2009. Thorax 64 J, Luque N, Ewig S, Menendez R,
23. Knaus WA, Draper EA, Wagner DP, (Suppl 3):iii1–iii55 Niederman MS, Torres A (2009) Severe
Zimmerman JE (1985) APACHE II a 29. Ambrosino N, Foglio K, Rubini F, Clini community-acquired pneumonia:
severity of disease classification system. E, Nava S, Vitacca M (1995) Non- Validation of the Infectious Diseases
Crit Care Med 13:818–829 invasive mechanical ventilation in acute Society of America/American Thoracic
24. Lim WS, van der Eerden MM, Laing R, respiratory failure due to chronic Society Guidelines to Predict an
Boersma WG, Karalus N, Town GI, obstructive pulmonary disease: Intensive Care Unit Admission. Clin
Lewis SA, Macfarlane JT (2003) correlates for success. Thorax Infect Dis 48:377–385
Defining community acquired 50:755–757 35. Aliberti S, Amir A, Peyrani P,
pneumonia severity on presentation to 30. Antonelli M, Conti G, Rocco M, Bufi Mirsaeidi M, Allen M, Moffett BK,
hospital: an international derivation and M, Deblasi RA, Vivino G, Gasparetto Myers J, Shaib F, Cirino M, Bordon J,
validation study. Thorax 58:377–382 A, Meduri GU (1998) A comparison of Blasi F, Ramirez JA (2008) Incidence,
25. Vincent JL, Moreno R, Takala J, noninvasive positive-pressure etiology, timing, and risk factors for
Willatts S, De Mendonca A, Bruining ventilation and conventional clinical failure in hospitalized patients
H, Reinhart CK, Suter PM, Thijs LG mechanical ventilation in patients with with community-acquired pneumonia.
(1996) The SOFA (Sepsis-related acute respiratory failure. N Engl J Med Chest 134:955–962
Organ Failure Assessment) score to 339:429–435 36. Lisboa T, Blot S, Waterer GW, Canalis
describe organ dysfunction/failure. On 31. Cheung MT, Yam LY, Lau CW, Ching E, de MD, Rodriguez A, Rello J (2009)
behalf of the Working Group on Sepsis- CK, Lee CH (2000) Use of non- Radiologic progression of pulmonary
Related Problems of the European invasive positive-pressure ventilation infiltrates predicts a worse prognosis in
Society of Intensive Care Medicine. for acute respiratory failure: prospective severe community-acquired pneumonia
Intensive Care Med 22:707–710 study. Hong Kong Med J 6:361–367 than bacteremia. Chest 135:165–172
26. Dellinger RP, Carlet JM, Masur H, 32. Rana S, Jenad H, Gay PC, Buck CF, 37. Goldhill DR, Sumner A (1998)
Gerlach H, Calandra T, Cohen J, Gea- Hubmayr RD, Gajic O (2006) Failure of Outcome of intensive care patients in a
Banacloche J, Keh D, Marshall JC, non-invasive ventilation in patients with group of British intensive care units.
Parker MM, Ramsay G, Zimmerman acute lung injury: observational cohort Crit Care Med 26:1337–1345
JL, Vincent JL, Levy MM (2004) study. Crit Care 10:R79 38. Goldhill DR, Withington PS (1996) The
Surviving sepsis campaign guidelines effect of casemix adjustment on
for management of severe sepsis and mortality as predicted by APACHE II.
septic shock. Crit Care Med Intensive Care Med 22:415–419
32:858–873