. 2014 Jun 3;9(6):e98580. doi: 10.1371/journal.pone.0098580

Table 3. Distribution of the outcomes, clinical evolution, and need for hospitalization, by subtype.

		Class
Outcome	Total	A	B	C	D	P^*
Need for	1537	507^cd	361^cd	404^abd	265^abc	<0·001^**
hospitalization	(61.8)	(54.3)	(52.9)	(71.9)	(85.8)
Death	59	9^cd	12^d	19^a	19^ab	<0·001^**
	(2.4)	(1.0)	(1.8)	(3.4)	(6.2)
ICU Admission	36	4^d	1^d	11^d	20^abc	<0·001^†
	(1.5)	(0.4)	(0.2)	(2.0)	(6.5)
Need for non-	98	12^d	6^d	18^d	62^abc	<0·001^‡
IMV	(3.9)	(1.3)	(0.9)	(3.2)	(20.1)
Other serious	547	236^d	182^cd	100^bd	29^abc	<0·001^§
complications	(22.0)	(25.3)	(26.7)	(17.8)	(9.4)
Total	2487	934	682	562	309
		(37.6)	(27.4)	(22.6)	(12.4)

Classes have been labeled in alphabetical order. Each cell shows the number of patients and column percentage. Abbreviations used in the text: intensive care unit (ICU), invasive mechanical ventilation (IMV).

*Statistically significant differences between groups are shown with superscripts represented by the code of the group (A, B, C and D).

**P-value based on the chi-squared test for difference in proportions between the groups using the whole sample (n = 2487).

^†

P-value based on the chi-squared test for difference in proportions of ICU admission using the subsample excluding the patients who died (n = 2428).

^‡

P-value based on the chi-squared test for difference in proportions of need for non-IMV for more than two days, using the subsample excluding the patients who had a poorer evolution (death, ICU admission) (n = 2392).

^§

P-value based on the chi-squared test for difference in proportions of other serious complications using the subsample excluding the patients who had a poorer evolution (death, ICU admission or need for non-IMV for more than two days) (n = 2294).