Included studies
Characteristics detailed in the included studies are shown in table 1. All the included studies are descriptive (17 retrospective(15)(18-20)(22-34), 1 prospective(35) and 4 series of cases.(36-39)
The studies were all in a single center. Most of the studies were from the USA (United States of America), (14) (18-19) (23) (26) (28) (30) (32-35) (37) 6 from Europe, (21- 22) (24-25) (31) (38) 2 from India, (29) (36) and 1 from Australia.
The total amount of patients in the studies was 1222, the age range between 0 and 21 years old. The most frequent underlying disease was acquired or congenital airway alteration, and chronic pulmonary disease.(22) (24-27) (29) (31) (33-36) (38-39) Few studies informed the severity of the disease in the included patients.(19) (22) (24-26) (31-33) (39)
None of the studies informed if the tracheostomy was performed by percutane or surgical technique. The indication of tracheostomy was informed on 76% of studies, most of the times by alterations in facial and upper airways.(17) (22-26) (31) (33) (39).
The duration of the mechanical ventilation support (MVS) was only measured in one study (around 19 days).(30) Even though the inclusion criteria were well-detailed in 15 studies(14) (18-19) (22) (24-30) (35-38) the exclusion criteria were only mentioned in 9 (14) (18-19) (29) (31) (33) (36-38). Decannulation protocols or algorithms were used in all studies with the exception of one.(34)Polysomnography (PSM) took place in half the studies(17-19) (25-28) (31) (33-34) (37) and only when \soutin was necessary in 3.(24) (29) (35)Radiography test was only mentioned in 2 protocols(30) (37) and endoscopy in most of them. In 9 studies(14) (17) (23-24) (27-28) (33) (35) (37) the indication to start the decannulation protocol was that patients were clinically stable and in 5 of them that the cause of tracheotomy was resolved.(22) (28) (30) (33) (35) In 11 studies(17) (23-26) (28-29) (33-34) (37) (39) the decrease of the cannula diameter was  the prior step to the its occlusion which was done in 14 studies.(19) (22-24) (26) (28-33) (36) (38-39) Only in three studies the cannula was changed, mostly for smaller cannulas.(22) (36) (38) Only at the beginning of two protocols there was an education plan for parents and intervening professionals.(30) (38) Decannulation criteria were presented in almost all the studies - 19 studies -  from which clinical and monitoring parameters such as partial oxygen saturation  (spo2) were selected by the majority as measures defining or non-defining decannulation. PSG was chosen by 8(17-19) (27-28) (31) (33) (37),cough effectivity only in 3 studies(24) (32-33) and endoscopy in 4.(14) (31-33)
Non-invasive ventilation (NIV) was used in 4 studies but after decannulation as a rescue maneuver.(25) (28-29) (34)The main outcome, in this case a successful decannulation, was only defined in 4 studies(18) (27) (35) (38) in two out of which the success was defined by the patient sustaining a stable decannulation for 6 months(27) (35) and in the other 2 studies for 2 years(18) or till the end of the tracing.(38) The post-decannulation observation period varied greatly among studies. The majority from 3 hours to 5 days(14) (17) (19) (23-24) (29) (31) (36) (39) or 3 to 6 months. (25) (27) (35)
Decannulation success ratio was over 70% in most studies - between 70 and 100% by protocol and in the first attempt - . As to PSG various authors concluded that in could be a good success predictor or, no less important, a parameter that could aid in decision making process when undertaking a decannulation.(17-19) (27-28) (33) (37)As to protocols usage it is mentioned in only two studies that it may aid to avoid a failure in decannulation but there are not any conclusions due to the nature of the designs.
Each study description is presented on the chart of included studies (chart 2).