OBJECTIVE: It is difficult to know when children hospitalized with pertussis can be safely discharged. We sought to identify clinical features of children hospitalized with pertussis that are associated with readmission. METHODS: A case series of 207 children hospitalized with pertussis was studied. The 33 children readmitted with pertussis were compared with the 174 who did not require readmission. RESULTS: Demographic characteristics and immunization status of the children with pertussis requiring readmission did not differ from the children who were not readmitted. Median duration of initial hospital stay was 4 days for both groups (P=.11). The children who were readmitted had more cyanotic episodes per day (0.8 vs 0.0 episodes, P=.03) and on greater proportion of hospital days (0.5 vs 0.1, P=.01). On the last day of admission, the children subsequently readmitted had more coughing episodes (4 vs 0, P < .001), and a larger proportion had a cyanotic episode (30% vs 10%, P=.003). The risk of readmission was increased in children who had > or =1 cyanotic episode per day (relative risk [RR]=2.5, 95% confidence interval [95% CI] 1.3-4.6); cyanosis on > or =50% of days (RR=2.6, 95% CI 1.4-4.8);> or =2 coughing paroxysms on the last hospital day (RR=2.4, 95% CI 1.3-4.4); or any cyanosis on the last day (RR=2.9, 95% CI 1.5-5.2). CONCLUSIONS: Paroxysmal cough and cyanosis are clinical signs that can be used in children hospitalized with pertussis to help decide when to discharge them from hospital.
OBJECTIVE: It is difficult to know when children hospitalized with pertussis can be safely discharged. We sought to identify clinical features of children hospitalized with pertussis that are associated with readmission. METHODS: A case series of 207 children hospitalized with pertussis was studied. The 33 children readmitted with pertussis were compared with the 174 who did not require readmission. RESULTS: Demographic characteristics and immunization status of the children with pertussis requiring readmission did not differ from the children who were not readmitted. Median duration of initial hospital stay was 4 days for both groups (P=.11). The children who were readmitted had more cyanotic episodes per day (0.8 vs 0.0 episodes, P=.03) and on greater proportion of hospital days (0.5 vs 0.1, P=.01). On the last day of admission, the children subsequently readmitted had more coughing episodes (4 vs 0, P < .001), and a larger proportion had a cyanotic episode (30% vs 10%, P=.003). The risk of readmission was increased in children who had > or =1 cyanotic episode per day (relative risk [RR]=2.5, 95% confidence interval [95% CI] 1.3-4.6); cyanosis on > or =50% of days (RR=2.6, 95% CI 1.4-4.8);> or =2 coughing paroxysms on the last hospital day (RR=2.4, 95% CI 1.3-4.4); or any cyanosis on the last day (RR=2.9, 95% CI 1.5-5.2). CONCLUSIONS:Paroxysmal cough and cyanosis are clinical signs that can be used in children hospitalized with pertussis to help decide when to discharge them from hospital.