Written by Clay Smith
It appears that there is no difference in treatment failure or mortality when low-risk febrile neutropenia patients are treated at home vs inpatient. But be very cautious about this.
Why does this matter?
Both the CISNE and MASCC scores have been retrospectively and prospectively derived and validated to determine which patients with fever and neutropenia are low risk and may avoid hospitalization. They have numerous caveats, and no RCTs have included them yet. So, they don’t show up in this meta-analysis. Despite this, ASCO and the IDSA have issued joint guidelines on risk stratification and outpatient treatment for select patients that combine both MASCC and CISNE scores into the algorithm. This is a review of several other RCTs in regard to outpatient management of febrile neutropenia.
Just schedule an appointment in the septic shock clinic…
Among patients with febrile neutropenia, 50-60% don’t have a life-threatening complication or fatal infection. This was a meta-analysis of 10 RCTs (6 adult/4 pediatric) comparing inpatient vs outpatient low-risk febrile neutropenia patients. “Low-risk” definitions were variable but usually meant patients would not: “a) need hospitalization, b) have focal or severe infection, c) have relapse of the disease, and d) be receiving intensive chemotherapy.” Overall, there was no difference in treatment failure or mortality when select patients were treated as outpatients. Most patients were watched in the hospital for 1-3 days. So, just 2 of 10 studies were true outpatient treatment with immediate discharge of low-risk patients. None used the MASCC or CISNE criteria, and there are no validated low-risk criteria for children. Overall quality of evidence was low to moderate, and most trials were small with wide confidence intervals. Larger, better quality studies are needed.
Here is my take. ASCO and IDSA consider outpatient management of febrile neutropenia to be a viable option. But be careful – the MASCC and CISNE scores don’t consider everything – not even close! This is a tough decision, with numerous medical and psychosocial caveats and should only be made with the patient’s oncologist and shared decision making with the patient.
See the full Cochrane Review of this topic.
See the new ASCO/IDSA Outpatient Management of Fever and Neutropenia Guidelines.
Outpatient Treatment for Low-Risk Febrile Neutropenia. Acad Emerg Med. 2019 Aug 17. doi: 10.1111/acem.13847. [Epub ahead of print]
Open in Read by QxMD
Reviewed by Thomas Davis
7 thoughts on “Outpatient Treatment of Febrile Neutropenia?”
I agree with some comments, but I have to disagree with some things.
In reality, CISNE was not developed to select patients for outpatient management, but for the opposite.
The aim was only to protect the most fragile patients from the recklessness of those who don’t read the trials carefully, and don’t reflect on them. For this reason, CISNE incorporates, in a natural way, the exclusion criteria for ambulatory management used in RCTs, did you noticed that?
Thus, CISNE could only be applied to patients with clinical stability, absence of complications and absence of serious infections, in a similar way to the criteria used in multiple clinical trials commented on in this blog. It is only a second safety step for the doctor to think twice about discharge.
Sadly, many people seem to read the articles superficially, and do not grasp these nuances. If used well, CISNE is therefore stricter than the eligibility criteria of clinical trials commented on here, which it complements.
This had a clear rationale based on the difficulty of generalizing results from RCTs in febrile neutropenia to daily practice, with patients often more fragile than in trials. Most RCTs on outpatient management declare some eligibility criteria, but in reality recruitment is usually far from being consecutive at each centre.
This implies that researchers tend to apply more rigorous de facto selection criteria than those reported in the articles (e.g., they recruit the fittest by far subgroup of all those meeting eligibility criteria). You may see that sample sizes are much lower that the real capacity of recruitment of most centers participating in these RCTs.
We actually found in a consecutive, prospective, data review series of >1000 patients that when these RCT-based criteria are applied consecutively to all, the subjects who meet them have a residual risk of unexpected serious complications of up to 13%. This is higher than reported in RCTs under the same circunstamces.
In addition, immunosuppression secondary to febrile neutropenia hinders the interpretation of signs and symptoms at the onset of infection, due to the inability to generate an inflammatory response. Therefore, clinical criteria are required but are not enough.
Eligibility criteria for RCTs are not entirely reliable in this setting. I assure you that if you trust these pragmatic RCT-based criteria alone, you will have two or three serious problems each year.
Hence, the need to strengthen these criteria.
So please read all model specifications correctly before applying them meaninglessly. CISNE should be used this way:
1) CISNE should never be applied alone, but in combination with rigorous clinical judgment and other criteria, to the extent that you believe that the latter is not sufficient.
2) The application of a precautionary principle implies that if CISNE indicates high risk, that patient should not ever be treated at home by this simple precautionary principle, at least until the apparent stability is confirmed as real.
3) If CISNE indicates low risk, the model cannot and does not attempt to influence decision making on outpatient or inpatient management.
Alberto Carmona Bayonas.
Great point on the exclusions. Clearly, the RCTs are very meticulously selecting patients for inclusion/exclusion. As I read the ASCO/IDSA statement, Table 4 lists an 41 additional clinical criteria (beyond MASCC and CISNE) that would preclude outpatient management. That is indeed rigorous clinical judgment! In the ED, we need to think through every organ system, use the decision tools available properly, and talk to the patient’s oncologist. And if there are any concerns, outpatient management is not a safe option. This comment is a helpful clarification and caution. I have to say, the concept of outpatient management for fever and neutropenia gives me a little heartburn… Many thanks!
I fully agree with you that randomized clinical trials are needed, and that the message should be prudence. Incredible as it may seem, the authors of MASCC state amazingly that this model can be used as a sole decision criterion, without taking clinical data into account, and even as opposed to clinical data (e.g., without Table 4 of the ASCO/IDSA guideline). For example, this is clearly explained here:
Talcott JA. Decision rules in a guideline: Allow the science to speak. J. Clin. Oncol. American Society of Clinical Oncology; 2018;2018.
This seems to me to be an absolute recklessness, not supported either by empirical data or by the basic principles of febrile syndrome management that we learn from university.
There is currently no evidence that the MASCC score can stratify clinically stable patients according to their risk of complications. For example, a retrospective single-center series examined patients with FN eligible for outpatient treatment based on exclusion criteria (e.g., clinical stability, normal organ function, absence of comorbidity, etc.) (Elting JCO 2008). A high readmission rate (21%) was found, revealing how the inability of immunocompromised patients to generate inflammatory reactions decreases the clinical expression of infection in the early stages. This justifies having models that support clinical data when such data alone are insufficient to detect severity. If, as suggested in the article, the MASCC score had been used instead of the clinical assessment, the results would not have improved, as MASCC classified 98% of outpatients as low-risk and was therefore unable to detect those at risk for complications. Interestingly, the authors come to the opposite conclusion, having misinterpreted their own data, possibly because of their illusions.
The most serious error of the MASCC score and, curiously, its strength (high positive predictive value), is that hypotension is both the most weighted predictor and the most common outcome; therefore, it literally predicts itself. It may seem like a joke, but it is literally so. No one says that people with shock should be admitted even without having to administer a model, which is shameful. Validations of the MASCC score actually prove that over and over again hypotension as covariable is associated with hypotension as a dependent variable (outcome) in all countries. However, after applying exclusion criteria (e.g., IDSA guideline table 4) in a clinically meaningful manner, then the sensitivity for detecting complications of the MASCC score drops to 33% (two out of three complications of this group are not detected). in fact, outpatient management studies based on MASCC consist of recording the MASCC score in the history and deciding with other criteria.
The authors report flow charts that reveal that the decision criteria are actually made on the basis of common sense, and not by a single MASCC score, but amazingly, articles tend to conclude that the merit is the MASCC model. For example, this and many others:
I really believe that this is a field of research things should and must be done much better.
Decision-making models should be applied to appropriate populations and, in the absence of other better validated through clinical trials, currently only CISNE helps to protect patients with clinical stability and febrile neutropenia from outpatient management.
CISNE is not just a model, but a multi-step decision making algorithm as outlined in the guide of the Spanish Society of Medical Oncology (SEOM 2018. Carmona-Bayonas A, et al. Clin Transl Oncol. 2019;21(1):75-86).
Initially, the clinician must select patients who do not meet exclusion criteria for outpatient management. This means that in order to apply CISNE it is essential that the patient has a situation of clinical stability, absence of complications and absence of serious infections.
The strengths of the CISNE model and especially its advantages with respect to MASSC are reflected in the 5 publications of the original series:
1) Carmona-Bayonas A, et al. J Clin Oncol. 2015;33(5):465-71;
2) Fonseca PJ, et al. Br J Cancer. 2016; 114(11):1191-8;
3) Carmona-Bayonas A, et al. Clin Transl Oncol. 2017;19(3):386-95;
4) Carmona-Bayonas A, et al. Clin Transl Oncol. 2017;19(9):1084-90;
5) Carmona-Bayonas A, et al. Eur J Intern Med. 2018;50:e33-e34.
CISNE goes further and seeks to be stricter than the exclusive use of these criteria, in the interests of safety. Therefore, CISNE is applied to patients who meet clinical stability criteria to avoid outpatient management if the model classifies them as having a high risk of complications. If CISNE indicates high risk, that patient should not be treated at home by a simple precautionary principle, at least until the apparent stability is confirmed as real.
In short, CISNE guides the intensification of follow-up in patients who appear stable but are at risk of complications and should only be applied when the use of empirical and strict exclusion criteria confirms clinical stability.
Dr Paula Jimenez-Fonseca, medical oncology, coordinator of the FINITE study – CISNE model
This is great feedback. It seems analogous to the PERC rule. Many forget that to be "PERC negative," it’s not just the 8 clinical variables. To even use PERC, gestalt must be low. CISNE sounds similar. To even consider using CISNE, the patient has to look well clinically, with no immediate exclusions to outpatient management. We want to get it right in this fragile patient population. And a decision like this means a call to someone like you! Many thanks for this clarification.