A total of 67 responses belonging to different clinics from 51 centers were received and analyzed. Figure 1 provides the distribution of participating centers across Italy. The majority of respondents were from the North of Italy (26, 38.8%), followed by the South (23, 34.3%, including the isles) and the Center (18, 26.9%). The number of patients with CVS cared among clinics was > 100 patients in 1 (1.5%), between 50 and 100 patients in 3 (4.5%), between 20 and 49 patients in 4 (6%), between 10 and 19 patients in 22 (32.8%) and less than 10 patients in 37 (55.2%).

Distribution of participating centers across Italy

Patients were followed-up in the gastroenterology outpatient clinic at 41 centers (61.2%), in the neurology outpatient clinic at 15 centers (22.4%), in the combined neuro-gastroenterology outpatient clinic at 9 centers (13.4%), in the headache outpatient clinic at 1 center (1.5%) and in a specific CVS outpatient clinic at 1 center (1.5%). Figure 2 depict the number of patients cared per specific clinic. No significant difference was found regarding number of patients cared according to specific clinic (p = 0.386, ns).

Number patients with cyclic vomiting syndrome cared per specific clinic

Rome IV diagnostic criteria were used by 41 clinics (61.2%), the International Classification of Headache Disorders (ICHD)-III beta criteria by 16 clinics (23.9%), the North American Society for Pediatric Gastroenterology, Hepatology & Nutrition (NASPGHAN) criteria by 8 clinics (11.9%), while 2 clinics (3%) applied both Rome IV and ICHD-III beta criteria (Fig. 3). Table 1 illustrate the criteria adopted according to specific outpatient clinic. A statistically significant difference was found between the adopted diagnostic criteria and the type of outpatient service (p = 0.002).

Cyclic vomiting syndrome diagnostic criteria generally adopted by respondents to the survey

The main organic causes of vomiting identified in patients with suspected CVS were gastrointestinal diseases in 36 clinics (53.7%), neurologic diseases in 33 (49.2%), metabolic and endocrinologic causes in 19 (28.3%), infections in 19 (28.3%), urologic causes in 3 (4.5%) and other causes in 3 (4.5%), which included genetic causes in 2 (3%) and appendicitis in 1 (1.5%). Supplementary Table 2 summarizes the main organic causes identified by clinicians according to specific outpatient clinic. No statistical difference was identified among the different clinic settings excepting for gastrointestinal diseases, which were more commonly identified by gastroenterologists (p = 0.006). Among responders to the open-ended question (27, 40.3%), pathologies identified as cause of vomiting were: epilepsy in 8 patients (Panayiotopoulos syndrome in 5), appendicitis in 5, food intolerance in 4, intestinal obstruction in 5, coeliac disease in 3, congenital cytomegalovirus infection in 3, volvulus in 2, pseudotumor cerebri in 2, inflammatory bowel disease in 2, esophagitis in 2, brain tumor in 2, gastritis in 1, Helicobacter Pylori infection in 1, post enteritis syndrome in 1, ion channel disease in 1, superior mesenteric artery syndrome in 1, GLUT-1 deficiency in 1, malrotation in 1, Chiari malformation in 1.

The majority of units (42 clinics, 62.7%) received a referral for suspected CVS after 3 to 5 acute episodes, whilst almost one third of centers (22, 32.8%) used to receive the referrals after more than 5 acute attacks. Three responders (4.5%) declared to receive referrals for suspected CVS only after 2 vomiting episodes. No statistical difference was found between number of attacked recorded at the time of referral and type of outpatient clinic (p = 0.86, ns).

The most common comorbidities recorded were headache by 63 clinics (94%), anxiety by 34 (50.7%), sleep disorders by 29 (43.3%), irritable bowel syndrome by 28 (41.8%); other comorbidities were recorded by 6 clinics (9%), and in particular neurodevelopmental delay by 2 (3%), constipation by 1 (1.5%), coeliac disease by 1 (1.5%), vertigo by 1 (1.5%) and eating disorder by 1 (1.5%). No comorbidities were recorded by 1 clinic (1.5%). Comorbidities recorded among patients with CVS according to specific outpatient clinic are summarized in Supplementary Table 3. No statistical difference was identified among the different clinic settings. Among responders to the open-ended question (26 clinics, 38.8%), the main comorbidities identified in patients with CVS were: headache in 107 patients, irritable bowel syndrome in 65, sleep disturbance in 51, anxiety in 22, migraine in 9, parasomnia in 4, autism in 1, coeliac disease in 1, abdominal pain in 4, delayed sleep induction in 1, vertigo in 1.

A family history of migraine was detected by 58 clinics (86.6%), functional GI disorders by 30 (44.8%), childhood periodic syndromes (including benign paroxysmal torticollis, benign paroxysmal vertigo, and abdominal migraine) by 20 (29.9%), cyclic vomiting syndrome by 10 (14.9%). Supplementary Table 4 depicts the family history of diseases detected among patients with CVS according to specific outpatient clinic. No statistical difference was identified among the different clinic settings excepting for childhood periodic syndromes (p = 0.046). Among responders to the open-ended question (22 clinics, 32.8%), a family history of migraine was identified in 206 patients, functional GI disorders in 37, CVS in 7, irritable bowel syndrome in 2, benign paroxysmal vertigo in 1.

Stress was the most common trigger encountered among patients and it was reported by 77.6% (52) of the responders, followed by sleep deprivation by 27 (40.3%), excessive excitement by 22 (32.8%), infections by 20 (30%), menstrual cycle by 9 (13.4%), physical exercise and foods by 6 (9%), and other factors by 3 (4.5%), including fever, sunlight exposure and car traveling. Supplementary Table 5 describes main triggers identified among patients according to specific outpatient clinic. No statistical difference was identified among the different clinic settings excepting for foods, who were more commonly identified by gastroenterologists and neuro-gastroenterologists (p = 0.019).

Among responders to the open-ended question (18 clinics, 26.9%), stress was a trigger factor in 105 patients, infection in 22, excessive excitement in 6, sleep deprivation 5 patients, menstrual cycle in 4.

Almost all survey responders (63, 94%) performed baseline testing toward identifying organic causes, including full blood count, C-reactive protein, electrolytes, glycemia, creatinine, uremia, aspartate aminotransferase, alanine aminotransferase, gamma-glutamyl transferase, amylase, lipase and urine analysis; 48 respondents (71.6%) performed blood gas analysis; 45 (67.2%) coeliac screening; 35 (52.2%) ammonia and lactic acid levels; 25 (37.3%) upper GI series; 23 (34.3%) ultrasound of the abdomen and pelvis; 24 (35.8%) electroencephalogram (EEG); 16 (23.9%) brain magnetic resonance imaging; 5 (7.5%) upper GI endoscopy with biopsies; 2 (3%) other investigations, such as urinary organic acids and serum amino acids, ophthalmology consultation and child neuropsychiatry consultations. Two respondents (3%) did not perform any investigation. Supplementary Table 6 summarizes screening tests performed according to specific outpatient clinic. No statistical difference was identified among the different clinic settings excepting for ammonia and lactic acid levels, which were more commonly performed by gastroenterologists (p = 0.015).

During the prodromal phase, the most widely used drug was the ondansetron (35 clinics, 52.2%), followed by sedatives, such as lorazepam and midazolam (11 clinics, 16.4%), sumatriptan (8, 11.9%), and aprepitant (6, 9%). Other drugs, such as anticonvulsants, analgesics, vitamin B6, steroids, and a combination of indomethacin, caffeine and prochlorperazine were also used (6 clinics, 9%). Twenty-three (34.3%) responders did not use any drug during the prodromal phase.

Ondansetron was the most commonly used medication during the emetic phase (56 clinics, 83.6%), together with supportive cares, such as decrease stimulation in a dark, quiet, private room (41, 61.2%), intravenous (IV) infusion of 10% glucose (35, 52.2%), or IV infusion of saline solution (14, 20.9%). Proton pump inhibitors were used during emetic phase by 17 clinics (25.4%), H2-receptor antagonists by 5 (7.5%), sedatives by 17 (26.9%), non-steroidal anti-inflammatory analgesics by 7 (10.4%), and other treatments, such as paracetamol, steroids, and oral rehydration, by 7 (9%).

Lifestyle changes and reassurance were promoted for the interictal period by 43 units (64.2%). Prophylactic pharmacotherapy commonly used was cyproheptadine by 34 (50.7%), pizotifen by 17 (25.4%), amitriptyline by 17 (25.4%), mitochondrial supplements by 10 (14.9%), anticonvulsants by 7 (10.4%), aprepitant 4 (6%), propranolol 4 (6%). Other treatments used by 8 units (11.9%) included flunarizine, 5-hydroxytryptophan, proton pump inhibitors, paracetamol, magnesium supplementation.

Supplementary Table 7 summarizes treatments applied in patients with CVS according to specific outpatient clinic. No statistical difference was encountered among the different clinic settings excepting for H2-receptor antagonists (p = 0.004) and sedatives (p = 0.03).

Regarding the most common long-term outcomes, progression to migraine was reported by 25 (37.3%), the persistence of symptoms with prolonged well-being phase by 21 responders (31.3%), complete resolution by 20 (29.9%), progression to other functional GI disorders by 6 (9%), and persistence of symptoms with the same characteristics by 4 (6%). Supplementary Table 8 depicts the long-term outcomes identified according to specific outpatient clinic. In particular, significant difference was reported among clinics regarding progression to migraine (p < 0.001) and persistence of symptoms with prolonged well-being phase (p = 0.024). No statistical correlation was found between the family history of disease or prophylactic therapy preferentially adopted and long-term outcomes.

Although to our knowledge the present study represents the most extensive data collection on current national practice concerning the diagnosis and management of patients with CVS, we have to acknowledge some limitations: the survey was designed to be rapid and easy to complete to involve a large number of centers, although the collected data lacked in precision; therefore, it does not guarantee the reliability of the results. Moreover, despite the large adhesion, we cannot be sure that all the centers that care for patients with CVS responded to the questionnaire. However, the aforementioned wide geographical distribution and the high response rate is likely to be a reliable national representation.