Table 4.

Random-effects model statistics: p-values for ANOVA tests (used to assess the effect of comparison group), p-values for Cochran’s Q tests and Higgins’ $I^2$ statistics (both used to assess heterogeneity)

Alpha-diversity index	Disease	Sample type	ANOVA p-value	Cochran’s Q p-value	$I^2$ statistic
Shannon	Asthma	BAL	0.81	0.01	$85\%$
		Sputum (induced)	0.98	$<0.01$	$92\%$
		UA	–	$<0.01$	$92\%$
		All	0.94	$<0.01$	$87\%$
	COPD	BAL	–	0.02	$80\%$
		All	0.59	$<0.01$	$87\%$
	CF	Sputum	0.04	0.07	$70\%$
		All	$<0.01$	0.16	$48\%$
	NCFB		–	–	–
	All		0.82	$<0.01$	$86\%$
Chao1	Asthma	UA	–	0.03	$78\%$
		All	0.24	0.03	$65\%$
	COPD		–	–	–
	CF		–	–	–
	NCFB		–	–	–
	All		0.09	$<0.01$	$68\%$
Simpson	Asthma	All	0.99	0.06	$71\%$
	COPD		–	–	–
	CF		–	–	–
	NCFB		–	–	–
	All		0.45	0.02	$59\%$

A line indicates that the test could not be performed (since only one comparison group, in the case of ANOVA, or insufficient number of studies, in the case of heterogeneity statistics)