Skip to main content

ROX index versus HACOR scale in predicting success and failure of high-flow nasal cannula in the emergency department for patients with acute hypoxemic respiratory failure: a prospective observational study



High-flow nasal cannula has been a promising initial respiratory support measure for patients with acute hypoxemic respiratory failure (AHRF) in the emergency department (ED). However, delayed detection of HFNC failure is associated with increased mortality. The ROX index is a tool that can help predict HFNC success. Nonetheless, its utility in ED patients is limited, and no studies have compared it with the HACOR scale, another tool that may be as accurate in predicting HFNC failure. Therefore, we aimed to compare the prognostic utility of the ROX index and the HACOR scale in emergency AHRF patients.


This prospective observational study was conducted at the ED of Siriraj Hospital, Thailand, between August 2018 and February 2020. Adult patients with AHRF requiring HFNC in the ED were included. The ROX index and the HACOR scale were measured at 1, 2, and 6 h after HFNC initiation. The primary outcome was HFNC success, defined as no intolerance or escalation towards mechanical ventilation or non-invasive ventilation within 48 h.


A total of 75 patients were enrolled; 52 (69.3%) had a successful treatment. The ROX index was higher in the success group, while the HACOR scale was lower at all timepoints. The ROX index yielded generally higher discrimination capacity based on the area under the receiver operating characteristic curve (AUROC) than the HACOR scale [AUROC at 1, 2, and 6 h = 0.815, 0.784, 0.853 for ROX in predicting HFNC success and 0.733, 0.690, and 0.764 for HACOR in predicting HFNC failure]. The ROX index measured at 6 h at the cut-point of 4.88 had 92.98% sensitivity, 61.11% specificity, 88.33% positive predictive value, and 73.33% negative predictive value with a diagnostic accuracy of 85.33%.


The ROX index had superior prognostic utility in predicting HFNC outcome (success/failure) compared to the HACOR scale in patients with AHRF in the ED setting. Moreover, it is less complex and more efficient to be employed at bedside. Therefore, the ROX index is a more appropriate tool to guide further management and potential escalation therapy for AHRF patients with HFNC therapy initiated in the ED.


Acute hypoxemic respiratory failure (AHRF) is a common and life-threatening presenting symptom to the emergency department (ED) [1]. Oxygen therapy is one of the most important initial treatment for AHRF [2]. It can be delivered via conventional technique through standard nasal cannula or facemask. However, such conventional oxygen delivering methods have several limitations including flow limitation, a varying fraction of inspired oxygen (FiO2), and insufficient heating and humidification [3, 4]. High-flow nasal cannula (HFNC) therapy is a novel device that can deliver the flow rate up to 60 L/min in adults and 100% heated and humidified oxygen via a large-bore nasal cannula [3, 4]. It has been proven to effectively improve oxygenation, decrease work of breathing, and lower the rate of escalation to invasive ventilation for patients with AHRF [5,6,7]. The efficacy of HFNC has also been demonstrated in the ED settings [8,9,10]. However, HFNC therapy needs to be used with caution, especially in critically ill patients, because delayed intubation in patients with failed HFNC treatment may increase mortality [11, 12]. Therefore, identifying factors and developing indices or scales that can predict the success or failure of HFNC are essential to reduce such delayed intubation and possibly prevent mortality. Although previous studies have demonstrated that severity scores in critically ill patients, such as sequential organ assessment (SOFA), simplified acute physiology II (SAP II), and acute physiology and chronic health evaluation II (APACHE II) score, were significant predictors of HFNC failure [13,14,15,16], these scores are complex and inconvenient to be employed as a bedside tool in the ED.

The ROX (Respiratory rate OXygenation) index, a ratio of oxygen saturation (SpO2)/FiO2 to respiratory rate, has been a known strong predictor of HFNC success [17, 18]. However, although it has been externally validated in many settings, most studies were still based in the intensive care units (ICU), with limited studies validating the utility of the ROX index in the ED, where patient characteristics and acuity may differ. On the other hand, the HACOR scale comprising of Heart rate, Acidosis, Consciousness (defined by the Glasgow Coma Scale [GCS] score), Oxygenation, and Respiratory rate (HACOR) is a score initially developed as a tool to predict non-invasive ventilation (NIV) failure [19]. Because NIV and HFNC both aim to prevent invasive ventilation, some studies have validated the HACOR scale for predicting HFNC failure, with most results showing favorable discrimination indices [20,21,22]. However, all studies were performed in the ICU, and only one study directly compared the two potential predictors [22]. Nonetheless, the study was conducted in a different patient population and setting, so their results may not be applicable to ED AHRF patients. We hypothesized that the HACOR scale might also have similar or superior prognostic utility for ED AHRF patients as the ROX index. Therefore, this study was conducted to assess the utility of the ROX index compared to the HACOR scale in predicting the success and failure of HFNC in AHRF patients in the ED.

Materials and methods

Study design and setting

This prospective observational study was conducted between August 12, 2018, and February 23, 2020, at the ED of Siriraj Hospital, the largest tertiary university hospital in Bangkok, Thailand, with over 2200 inpatient beds. The ED accommodates over 18,000 annual high-acuity visits triaged as level 1 or 2 based on the Emergency Severity Index (ESI) criteria [23]. The study was approved by Siriraj Institutional Review Board (certificate no. 224/2018). Written informed consent was obtained from all participants or their next of kin by the study investigators prior to the study inclusion. Standard treatment with NIV or HFNC was given to those who declined to participate at the attending physician’s discretion.


Adult patients over 18 years of age diagnosed with AHRF and determined by the attending ED physicians to require supportive oxygen therapy via HFNC were included. AHRF requiring HFNC was determined when all of the following criteria were met; a respiratory rate of more than 24 breaths/min, the presence of accessory muscle use, room air SpO2 of equal to or less than 90%, and the need for oxygen supplement via face mask > 9 L/min to maintain SpO2 > 92% [24,25,26]. Exclusion criteria were immediate cardiac or respiratory arrest or respiratory failure (respiratory rate > 35 breaths/min or SpO2 < 90% despite oxygen supplement via facemask > 9 L/min) requiring immediate mechanical ventilation, GCS < 12, hemodynamic instability (systolic blood pressure < 90 mmHg or mean arterial pressure < 65 mmHg), severe respiratory acidosis (pH < 7.3 and arterial pressure of carbon dioxide > 50 mmHg) or suspected hypercapnic respiratory failure, pneumothorax, and do-not-intubate status.

Study process and data collection

Patients visiting the ED with AHRF were consecutively assessed for eligibility. After ED arrival, eligible patients first received oxygen supplement via conventional technique and standard medical treatment as considered appropriate by the attending ED physician. They were then re-evaluated at 10 min and recruited if they were still eligible and did not meet any exclusion criteria. After informed consent was obtained, HFNC (AIRVO®2; Fisher & Paykel, Auckland, New Zealand) was applied at an initial flow rate of 30 L/min and FiO2 of 0.5. The flow could be adjusted up to 60 L/min as tolerated by the participants with FiO2 adjusted to maintain SpO2 of more than 94%. HFNC was terminated, and respiratory support was escalated to either NIV or mechanical ventilation if one of the following termination criteria, adopted from the British Thoracic Society’s guideline [27], was met; worsening dyspnea or hypoxemia (respiratory rate > 40 breaths/min or SpO2 < 90% despite maximum FiO2), worsening hypercapnia (increased PaCO2 or pH < 7.30 despite maximum standard medical care), deterioration of consciousness (GCS < 12 or decrease > 2 from baseline), severe hemodynamic instability (norepinephrine > 0.1 μg/kg/min or dopamine > 20 mcg/kg/min), and at the physician’s discretion. Other treatment was given as appropriate under the discretion of the treating physician.

Demographic data, including age, sex, and underlying diseases, were recorded. Respiratory and physiologic parameters (respiratory rate, signs of respiratory distress, pulse rate, blood pressure, SpO2, and GCS) and arterial gas results were recorded at initial ED presentation, 0, 1, 2, and 6 h after the initiation of HFNC. Participants’ diagnoses, management, and outcomes were also recorded.

Study parameters and outcomes

At 1, 2, and 6 h after HFNC initiation, two parameters were assessed for their utility in predicting HFNC success: the ROX index and SpO2/FiO2 (SF) ratio. The ROX index was calculated from the ratio of SF to respiratory rate [17]. At the same timepoints, two parameters were evaluated for their predictive ability of HFNC failure: respiratory rate and the HACOR scale. The HACOR scale (0–25 points) consists of five components with weighted score points [19]. The components are pulse rate (0–1 point), pH (0–4 points), GCS (0–10 points), PaO2/FiO2 (0–6 points), and respiratory rate (0–4 points).

Successful HFNC was defined as no intolerance requiring equipment removal and no escalation to invasive or non-invasive ventilation within 48 h after HFNC application. The primary aim of the study was to assess the predictive accuracy of the ROX index and the HACOR scale in predicting HFNC success and failure, respectively. In addition, the components of the ROX index, namely SF ratio and respiratory rate, were also individually evaluated for their prognostic utility. This was performed to explore the components contributing the most to the prognostic utility of the ROX index. Moreover, we also explored the potential prognostic utility of pulse rate as an exploratory analysis since it is also one of the components of the HACOR scale, which is a known strong predictor of HFNC failure [28, 29].

Statistical analyses

The initial sample size was calculated based on an expected sensitivity and specificity of 70% and 72%, respectively, for the ROX index at the recommended cut-point of 4.88 in predicting successful HFNC [17]. We estimated that the ROX index would have 70 ± 9% accuracy in determining HFNC success; thus, a sample size of 100 participants was required. As for the HACOR scale, we estimated that the sample size required would be equal to or lower than those for the ROX index, as we hypothesized that the HACOR scale would yield similar accuracy with a slightly higher standard error than that of the ROX index. However, we had to discontinue the trial prematurely because of the COVID-19 outbreak causing resource limitations after 75 participants were recruited.

Descriptive statistics were employed to present patient demographics, physiologic variables, and outcomes. These characteristics were compared between HFNC success and failure groups using the independent t-test or the Mann–Whitney U test for normally distributed and non-normally distributed continuous data, respectively, and the chi-squared or Fisher’s exact test for categorical data. The accuracy of the parameters in predicting the success or failure of HFNC was analyzed by using the receiver operating characteristic (ROC) curves and the area under the curves (AUROC) and its 95% confidence interval (CI). We further explored the AUROC for the subgroup with and without pneumonia as an exploratory analysis. Sensitivity, specificity, positive and negative predictive value (PPV and NPV), positive and negative likelihood ratios (LR + and LR −), and accuracy of the ROX index at the recommended cut-point (4.88) and the optimal cut-point from the data (the point from which the highest sensitivity and specificity were derived) were analyzed and presented. Moreover, we performed the log-rank test of the survival function contrasting the ROX index at the cut-point of 4.88 at all timepoints with HFNC failure as the outcome. We also performed univariate and multivariate logistic regression analyses to identify independent predictors of HFNC success. Age, sex, and underlying respiratory diseases were determined a priori as potential associating variables with HFNC success to be adjusted for in the multivariate logistic regression models. Only one parameter value at one timepoint was evaluated in each multivariate regression model to avoid multicollinearity.

A p-value of less than 0.05 was considered statistically significant. All statistical analyses were performed using SPSS 18 (SPSS Inc., Chicago, Illinois) except for sensitivity and specificity, LR + , LR − , NPV, PPV, and accuracy, which were calculated using MedCalc for Windows version 19 (MedCalc statistical software, Mariakerke, Belgium).


Between August 12, 2018, and February 23, 2020, a total of 587 patients visited the ED with AHRF. Of these, 258 did not require HFNC, 112 required immediate mechanical ventilation, 73 had do-not-intubate status, 68 had depressed GCS, 32 had either severe acidosis or suspected hypercapnic respiratory failure, and 17 had unstable hemodynamic status. Consequently, a total of 75 patients received HFNC and were included (Fig. 1). Their characteristics are shown in Table 1. Of all included patients, 26 were male (34.5%), and their mean age was 69.12 ± 17.69 years. Twenty-three patients (30.7%) had met at least one criteria of HFNC failure, while 52 (69.3%) were successfully treated with HFNC. Baseline demographics and physiologic variables were similar between those with HFNC success and failure status except for higher initial pulse rate and respiratory rate in the failure group (p = 0.003 and 0.026, respectively). Primary diagnoses were similar despite a trend of higher pneumonia rate in the failure group and a higher proportion of cardiogenic pulmonary edema patients in the success group. Hospital length of stay and mortality rate was significantly higher in the failure group (p = 0.047 and 0.001, respectively).

Fig. 1
figure 1

Study flow. AHRF, acute hypoxemic respiratory failure; ED, Emergency Department; HFNC, high-flow nasal cannula

Table 1 Patient characteristics by high-flow nasal cannula success status

All the four potential predictors of HFNC outcome were mostly different compared between the success and the failure groups at all timepoints. The ROX index was generally increasing over time and was significantly lower in the failure than the success group at all timepoints (Table 2). On the other hand, the HACOR scale decreased over time and was significantly higher in the failure groups (Table 2). SF ratio changed and differed between the groups in the same pattern as the ROX index, while respiratory rate followed the pattern of the HACOR scale. Nonetheless, the between-group differences were more prominent at later timepoints for these two parameters, as opposed to the HACOR scale, whose difference was most notable on the first assessment (Table 2).

Table 2 Descriptive statistics of potential predictors of success and failure of high-flow nasal cannula

The ROC curves of all parameters are shown in Fig. 2. The ROX index is the best predictor of HFNC outcome with the highest discriminating capacity for HFNC success based on AUROC (AUROC 0.815, 0.784, 0.853 at 1, 2, and 6 h, respectively) (Table 3). In contrast, the HACOR scale showed an even lower discriminating ability to differentiate HFNC failure than respiratory rate with lower AUROC at all three timepoints (AUROC 0.733, 0.690, 0.702 at 1, 2, and 6 h, respectively). Also, the ROX index yielded the highest discriminating capacity than the other parameters in both the subgroup with and without pneumonia (Table S1). Moreover, the ROX index was a better predictor of HFNC success than the HACOR scale based on the logistic regression analyses. The ROX indices at all three timepoints were significant and independent predictors of HFNC success in all univariate and multivariate logistic regression models. In contrast, the HACOR scale failed to yield statistically significant associations with the outcome in some models, also with a generally lower strength of association in all models (Table 4). As for pulse rate, although it was generally higher in the failure group at all timepoints, its association with HFNC success was not as strong as those of respiratory rate and the ROX index (Tables 2 and 4).

Fig. 2
figure 2

Receiver operating characteristic curves of potential predictors of high-flow nasal cannula success (A) and failure (B) at different timepoints. SF, pulse oximetry/fraction of inspired oxygen ratio; RR, respiratory rate

Table 3 Area under the receiver operating characteristic curve (AUROC) of the parameters at 1, 2, and 6 h in predicting high-flow nasal cannula success or failure
Table 4 Univariate and multivariate analyses of potential factors predicting high-flow nasal cannula success

We further evaluated the diagnostic accuracy indices of the ROX index, the best predictive parameter, based on the recommended cut-point of 4.88 and the optimal cut-point from our data of 4.75. The results are shown in Table 5. The highest sensitivity (94.23%) and NPV (76.92%) were achieved with the ROX index > 4.75 measured at 2 h, while the highest specificity (61.11%) and PPV (88.33%) were seen when analyzing the 6-h ROX index at either cut-points. Similarly, the ROX index at 6 h had the highest accuracy in predicting HFNC success. Additionally, patients with the ROX index of less than 4.88 had significantly lower survival function for HFNC failure based on the log-rank test compared to those with higher ROX at all three timepoints (log-rank p-value = 0.001, < 0.001, and < 0.0001 at 1, 2, and 6 h, respectively; Fig. 3).

Table 5 Diagnostic indices of the ROX index at the recommended (4.88) and optimal (4.75) cut-points
Fig. 3
figure 3

Survival function for high-flow nasal cannula failure of patients with the ROX index lower than 4.88 (blue line) compared to those with the index equal to or higher than 4.88 (red line) at 1 h (A), 2 h (B), and 6 h (C). The p-value for log rank test comparing survival functions between the groups at the ROX index cut-point of 4.88 = 0.001 (A) and < 0.001 (B and C)


To the best of our knowledge, this study was the first to directly compare the prognostic utility of the ROX index and the HACOR scale for undifferentiated AHRF in the ED setting. It was found that the ROX index was the best predictor of HFNC outcome since it could yield the highest discriminating ability for HFNC success. The index performed better once evaluated at 6 h post-HFNC application with the highest AUROC and most diagnostic indices higher than those measured at 1 or 2 h. In contrast, the HACOR scale was not a good predictor of HFNC failure, with the lowest discriminating ability compared to the other parameters.

HFNC has been increasingly used as the initial respiratory and oxygenation support measure in the ED. It has been shown to provide many physiologic benefits and may reduce adverse outcomes, such as mechanical ventilation rate [7,8,9, 30]. However, patients who fail HFNC treatment may still end up requiring invasive ventilation and a delay in such a process may lead to mortality [11, 12]. Therefore, appropriate and close observation with effective and efficient tools is mandatory to prevent these unfavorable consequences. The ROX index, one of the most widely validated prognostic indices for HFNC, was proposed for such purpose [17]. It has been proved as an accurate tool in predicting HFNC success in many studies, most of which were conducted in patients with pneumonia in the ICU setting [17, 18], whereas studies in the ED, whereby patient characteristics may be different and the cause of AHRF is usually unknown, are limited. Consequently, the present study has added to the current body of evidence that the ROX index is also an accurate predictor of HFNC success in the ED setting. Its AUROC was even slightly higher than those reported in some other external validation studies, and it could also yield favorable and comparable diagnostic accuracy indices at either of the cut-points studied [17, 18, 22]. The recommended cut-point of 4.88 from the literature [17] was feasible in our setting with almost as high diagnostic accuracy indices as the optimal cut-point from the data and also with highly significant log-rank tests. Additionally, we found that its prognostic utility seems to increase with time of HFNC treatment, similar to the ROX development study [18], which may have important clinical implications if the index measured at earlier timepoints are to be employed for clinical use in the ED as their lower predictive ability should be kept under consideration.

As for the HACOR scale, the present study found that it was the worst predictor of HFNC outcome among all the parameters studied. Its AUROC was largely smaller than those of the ROX index at all timepoints. One of the reasons why the HACOR scale was not a good predictor of HFNC failure was its relatively low weight anchored towards pulse rate, another known strong predictor of HFNC outcomes [28, 29]. However, this result was discordant with the only study that performed a pairwise comparison between the two parameters [22]. Despite similar AUROC of the HACOR scale between the present study and that previous study by Valencia et al. [22], they reported comparable discriminating capacity of the two parameters while we found the ROX index to be highly superior. This discordance could have most likely been due to different study population since they only enrolled patients with acute respiratory failure due to COVID-19 pneumonia in the ICU [22]. Another study by Magdy et al. [20] also evaluated the utility of the HACOR scale in predicting HFNC failure in the ICU and found that the HACOR scale could yield very high AUROC and diagnostic indices, contrary to our findings. The difference could have also been due to different study population. Most of the patients included in their study had pneumonia, while a largely higher proportion of our participants had cardiogenic pulmonary edema. In fact, in our subgroup analysis, we found that the AUROC of the ROX and the HACOR scale were more comparable in the pneumonia subgroup, which was concordant with those previous studies conducted in a similar patient population. Regardless, it would be interesting to explore the prognostic utility of these parameters in populations other than pneumonia patients, such as those with cardiogenic pulmonary edema, the most common second leading cause of AHRF [31]. Also, further studies should focus more on the potential benefit of HFNC in this population since we found a trend towards higher HFNC success in patients with AHRF than in those with pneumonia.

From the present study, it is appropriate to conclude that the ROX index is superior to the HACOR scale in predicting HFNC outcomes for undifferentiated AHRF in the ED setting not only because of the overall statistical indices but also due to the fact that the ROX index is easier to calculate at bedside and does not require arterial gas analysis, thereby being more efficient to use in the ED.

The present study also had limitations. First, it was a single-center study, possibly limiting the generalizability of the study results. Second, the sample size was smaller than intended since the study had to be discontinued due to the first COVID-19 outbreak, causing most analyses to be underpowered. The small sample size resulted in wide standard errors of the AUROCs, causing the differences to be most likely non-statistically significant. Despite such an issue, we could still demonstrate a substantial difference between the point estimates of the AUROC of the two main study parameters. Therefore, we can be quite certain of the relative difference in discrimination capacity between the two parameters, although the exact effect estimates may not be accurate due to random error. Third, despite the studying having a clearly defined termination and escalation criteria that did not include the ROX index, there could still have been some physicians who determined their treatment decisions based on the ROX index, a known widely used parameter, thus possibly causing a falsely high association and predictive ability for the ROX index. Fourth, although it would have been interesting had the parameters been followed longer than 6 h with the increasing trend of prognostic utility observed, the relevance of such results to the ED setting may be limited. Nevertheless, future multicenter studies with an adequate sample size including patients of various etiologies of AHRF should be performed to confirm the present study’s results and explore further prognostic utility of these parameters.


The ROX index was better than the HACOR scale at predicting HFNC outcome (success/failure) in patients with AHRF of undifferentiated origin in the ED setting. It has superior prognostic utility with less complexity and higher efficiency to be employed at bedside compared to the HACOR scale. Therefore, the ROX index is a more appropriate tool to guide further management and potential escalation therapy for AHRF patients with HFNC therapy initiated early in the ED.

Availability of data and materials

The dataset is not available but can be requested from the corresponding author.



Acute hypoxemic respiratory failure


Emergency department


High-flow nasal cannula

FiO2 :

Fraction of inspired oxygen


Intensive care unit


Sequential organ assessment


Simplified acute physiology II


Acute physiology and chronic health evaluation II


Respiratory rate Oxygenation

SpO2 :

Oxygen saturation


Heart rate, acidosis, consciousness, oxygenation, and respiratory rate


Glasgow Coma Scale


Non-invasive ventilation


Receiver operating characteristic


Area under the curve of the receiver operator characteristic curves

LR + :

Positive likelihood ratio


Negative likelihood ration


Negative predictive value


Positive predictive value


  1. Nee PA, Al-Jubouri MA, Gray AJ, O’Donnell C, Strong D. Critical care in the emergency department: acute respiratory failure. Emerg Med J. 2011;28(2):94–7.

    Article  Google Scholar 

  2. O’Driscoll BR, Howard LS, Davison AG. British Thoracic Society. BTS guideline for emergency oxygen use in adult patients. Thorax. 2008;63(Suppl 6):vi1-68.

    Google Scholar 

  3. Nishimura M. High-flow nasal cannula oxygen therapy in adults. J Intensive Care. 2015;3(1):15.

    Article  Google Scholar 

  4. Spoletini G, Alotaibi M, Blasi F, Hill NS. Heated humidified high-flow nasal oxygen in adults: mechanisms of action and clinical implications. Chest. 2015;148(1):253–61.

    Article  Google Scholar 

  5. Roca O, Riera J, Torres F, Masclans JR. High-flow oxygen therapy in acute respiratory failure. Respir Care. 2010;55(4):408–13.

    Google Scholar 

  6. Sztrymf B, Messika J, Bertrand F, Hurel D, Leon R, Dreyfuss D, et al. Beneficial effects of humidified high flow nasal oxygen in critical care patients: a prospective pilot study. Intensive Care Med. 2011;37(11):1780–6.

    Article  CAS  Google Scholar 

  7. Bell N, Hutchinson CL, Green TC, Rogan E, Bein KJ, Dinh MM. Randomised control trial of humidified high flow nasal cannulae versus standard oxygen in the emergency department. Emerg Med Australas. 2015;27(6):537–41.

    Article  Google Scholar 

  8. Makdee O, Monsomboon A, Surabenjawong U, Praphruetkit N, Chaisirin W, Chakorn T, et al. High-flow nasal cannula versus conventional oxygen therapy in emergency department patients with cardiogenic pulmonary edema: a randomized controlled trial. Ann Emerg Med. 2017;70(4):465-472.e2.

    Article  Google Scholar 

  9. Ruangsomboon O, Limsuwat C, Praphruetkit N, Monsomboon A, Chakorn T. Nasal high-flow oxygen versus conventional oxygen therapy for acute severe asthma patients: a pilot randomized controlled trial. Acad Emerg Med. 2021;28(5):530-541.

  10. Ruangsomboon O, Dorongthom T, Chakorn T, Monsomboon A, Praphruetkit N, Limsuwat C, et al. High-flow nasal cannula versus conventional oxygen therapy in relieving dyspnea in emergency palliative patients with do-not-intubate status: a randomized crossover study. Ann Emerg Med. 2020;75(5):615–26.

    Article  Google Scholar 

  11. Kang BJ, Koh Y, Lim CM, Huh JW, Baek S, Han M, et al. Failure of high-flow nasal cannula therapy may delay intubation and increase mortality. Intensive Care Med. 2015;41(4):623–32.

    Article  Google Scholar 

  12. Saillard C, Lambert J, Tramier M, Chow-Chine L, Bisbal M, Servan L, et al. High-flow nasal cannula failure in critically ill cancer patients with acute respiratory failure: Moving from avoiding intubation to avoiding delayed intubation. PLoS ONE. 2022;17(6): e0270138.

    Article  CAS  Google Scholar 

  13. Roca O, de Acilu MG, Caralt B, Sacanell J, Masclans JR, ICU collaborators. Humidified high flow nasal cannula supportive therapy improves outcomes in lung transplant recipients readmitted to the intensive care unit because of acute respiratory failure. Transplantation. 2015;99(5):1092–8.

    Article  Google Scholar 

  14. Hyun Cho W, Ju Yeo H, Hoon Yoon S, Lee S, SooJeon D, Seong Kim Y, et al. High-flow nasal cannula therapy for acute hypoxemic respiratory failure in adults: a retrospective analysis. Intern Med. 2015;54(18):2307–13.

    Article  Google Scholar 

  15. Messika J, Ben Ahmed K, Gaudry S, Miguel-Montanes R, Rafat C, Sztrymf B, et al. Use of high-flow nasal cannula oxygen therapy in subjects with ARDS: a 1-year observational study. Respir Care. 2015;60(2):162–9.

    Article  Google Scholar 

  16. Rello J, Pérez M, Roca O, Poulakou G, Souto J, Laborda C, et al. High-flow nasal therapy in adults with severe acute respiratory infection: a cohort study in patients with 2009 influenza A/H1N1v. J Crit Care. 2012;27(5):434–9.

    Article  Google Scholar 

  17. Roca O, Messika J, Caralt B, García-de-Acilu M, Sztrymf B, Ricard JD, et al. Predicting success of high-flow nasal cannula in pneumonia patients with hypoxemic respiratory failure: The utility of the ROX index. J Crit Care. 2016;1(35):200–5.

    Article  Google Scholar 

  18. Roca O, Caralt B, Messika J, Samper M, Sztrymf B, Hernández G, et al. An index combining respiratory rate and oxygenation to predict outcome of nasal high-flow therapy. Am J Respir Crit Care Med. 2018;199(11):1368–76.

    Article  Google Scholar 

  19. Duan J, Han X, Bai L, Zhou L, Huang S. Assessment of heart rate, acidosis, consciousness, oxygenation, and respiratory rate to predict noninvasive ventilation failure in hypoxemic patients. Intensive Care Med. 2017;43(2):192–9.

    Article  CAS  Google Scholar 

  20. Magdy DM, Metwally A. The utility of HACOR score in predicting failure of high-flow nasal oxygen in acute hypoxemic respiratory failure. Adv Respir Med. 2021;89(1):23–9.

    Article  Google Scholar 

  21. Salvador-Ibarra IJ, Dávila AP. Prediction of failure of the high-flow nasal tips with the HACOR scale, in acute respiratory insufficiency in patients with SARS-CoV-2. Med Crit. 2021;35(5):250–5.

    Google Scholar 

  22. Valencia CF, Lucero OD, Castro OC, Sanko AA, Olejua PA. Comparison of ROX and HACOR scales to predict high-flow nasal cannula failure in patients with SARS-CoV-2 pneumonia. Sci Rep. 2021;11(1):22559.

    Article  CAS  Google Scholar 

  23. Gilboy N, Tanabe P, Travers DA, Rosenau AM, Eitel DR. Emergency Severity Index, version 4: implementation handbook. Rockville: AHRQ Publication; 2005.

    Google Scholar 

  24. O’Driscoll BR, Howard LS, Earis J, Mak V. British Thoracic Society Guideline for oxygen use in adults in healthcare and emergency settings. BMJ Open Respir Res. 2017;4(1): e000170.

    Article  Google Scholar 

  25. Ponikowski P, Voors AA, Anker SD, Bueno H, Cleland JGF, Coats AJS, et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: The Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC) Developed with the special contribution of the Heart Failure Association (HFA) of the ESC. Eur Heart J. 2016;37(27):2129–200.

    Article  Google Scholar 

  26. Drake MG. High-flow nasal cannula oxygen in adults: an evidence-based assessment. Annals ATS. 2018;15(2):145–55.

    Article  Google Scholar 

  27. British Thoracic Society Standards of Care Committee. Non-invasive ventilation in acute respiratory failure. Thorax. 2002;57(3):192–211.

    Article  Google Scholar 

  28. Goh KJ, Chai HZ, Ong TH, Sewa DW, Phua GC, Tan QL. Early prediction of high flow nasal cannula therapy outcomes using a modified ROX index incorporating heart rate. J Intensive Care. 2020;8(1):41.

    Article  Google Scholar 

  29. Karim HMR, Esquinas AM. Success or failure of high-flow nasal oxygen therapy: The ROX Index is good, but a Modified ROX Index may be better. Am J Respir Crit Care Med. 2019;200(1):116–7.

    Article  Google Scholar 

  30. Frat JP, Thille AW, Mercat A, Girault C, Ragot S, Perbet S, et al. High-flow oxygen through nasal cannula in acute hypoxemic respiratory failure. N Engl J Med. 2015;372(23):2185–96.

    Article  CAS  Google Scholar 

  31. Ruangsomboon O, Praphruetkit N, Monsomboon A. Parallel-group, randomised, controlled, non-inferiority trial of high-flow nasal cannula versus non-invasive ventilation for emergency patients with acute cardiogenic pulmonary oedema: study protocol. BMJ Open. 2022;12(7): e052761.

    Article  Google Scholar 

Download references


Not applicable.


The study received no funding.

Author information

Authors and Affiliations



NP and OR conceived the study, designed the trial, and supervised the conduct of the trial. NP, NB, and AM completed the data collection. NP, NB, and OR managed the data. OR analyzed the data. NP and OR drafted the article. OR takes responsibility for the paper as a whole. The author(s) read and approved the final manuscript.

Corresponding author

Correspondence to Onlak Ruangsomboon.

Ethics declarations

Ethics approval and consent to participate

The study was approved by Siriraj Institutional Review Board (certificate of approval Si 224/2018). All participants or their next of kin provided written informed consent prior to the study inclusion. The study was conducted in accordance with the Declaration of Helsinki and adhered to Good Clinical Practice guidelines.

Consent for publication

Not applicable.

Competing interests

Fisher & Paykel Healthcare supported all HFNC materials for the study but did not involve in handling of any study data or manuscript preparation. The authors declare that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: Table S1.

Area under the receiver operating characteristics curve (AUROC) of the parameters at 1, 2, and 6 hours in predicting high-flow nasal cannula success or failure in a subgroup with pneumonia and without pneumonia.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Praphruetkit, N., Boonchana, N., Monsomboon, A. et al. ROX index versus HACOR scale in predicting success and failure of high-flow nasal cannula in the emergency department for patients with acute hypoxemic respiratory failure: a prospective observational study. Int J Emerg Med 16, 3 (2023).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: