Table 4.

Associations Between Baseline Variables and 10-Week Mortality

Variable	Category	No.	10-wk Mortality	OR (95% CI), Univariable	P Value	AOR (95% CI), Multivariablea,b	P Value
Age	<50 y	454	33% (148)	1	.014	1	.009
	≥50 y	24	58% (14)	2.9 (1.2–6.8)		4.0 (1.4–11.4)
Sex	Female	231	34% (79)	1	.815
	Male	250	34% (84)	1.0 (.6–1.4)
Seizures	No	389	33% (127)	1	.912
	Yes	91	37% (34)	1.0 (.6–1.6)
Mental status	Normal	366	25% (90)	1	<.001	1	<.001
	Abnormal	117	62% (73)	5.2 (3.3–8.3)		2.8 (1.6–4.7)
Weight	<50 kg	173	39% (68)	1	.003	1	.004
	≥50 kg	277	25% (68)	0.5 (.3–.8)		0.6 (.4–1.0)
Pulse	≤100 bpm	389	31% (121)	1	.010
	>100 bpm	86	45% (39)	1.9 (1.2–3.1)
Respiratory rate	≤20 bpm	363	30% (110)	1	.006
	>20 bpm	84	45% (38)	2.0 (1.2–3.4)
CD4 cell count	<25 cells/µL	226	35% (80)	1	.03c	1	.781
	25–49 cells/µL	102	30% (30)	0.7 (.4–1.2)		0.8 (.5–1.4)
	50–99 cells/µL	73	23% (17)	0.6 (.3–1.0)		0.8 (.4–1.5)
	≥100 cells/µL	39	23% (9)	0.5 (.2–1.2)		0.7 (.3–1.9)
Hemoglobin	≥7.5 g/dL	423	31% (133)	1	.008	1	.02
	<7.5 g/dL	27	56% (15)	3.0 (1.3–6.4)		3.0 (1.2–7.4)
White blood cell count	≤10 × 109/L	377	30% (114)	1	.001	1	.02
	>10 × 109/L	21	63% (13)	4.7 (1.8–12.2)		4.0 (1.3–12.6)
CSF opening pressure	≤25 cm CSF	213	39% (83)	1	.009	1	.002
	>25 cm CSF	223	30% (66)	0.6 (.4–.9)		0.4 (.3–.7)
CSF white cell count	≤20 × 106/L	268	35% (93)	1	.461
	>20 × 106/L	179	31% (55)	0.9 (.6–1.3)
QCC	1st tertile	161	24% (38)	1	<.001	1.3 (1.1–1.7)	.007
	2nd tertile	161	32% (52)	1.5 (.9–2.4)		(per log10 CFU/mL increase)d
	3rd tertile	158	46% (72)	2.8 (1.7–4.5)
Treatment	Fluconazole	99	53% (52)	1	<.001	1	.02
	Amphotericin	385	29% (111)	0.4 (.2–.6)		0.5 (.3–.9)

Abbreviations: AOR, adjusted odds ratio; CI, confidence interval; CFU, colony-forming units; CSF, cerebrospinal fluid; OR, odds ratio; QCC, quantitative cryptococcal culture.

ORs and 95% CIs for both univariable and multivariable associations are adjusted for clustering by study using a random-effects term for “study” in a hierarchical mixed-effects logistic regression model. There was very little evidence for significant clustering by study in either the 2-week or 10-week model (likelihood-ratio test of ρ = 0, P = .498 at both 2 and 10 weeks).

Numbers of patients included in each analysis are indicated in the table. A complete records analysis was performed rather than multiple imputation as there were relatively few missing data points in the key exposure and outcome variables, and missing variables in important exposure variables such as CD4 cell count were thought to be missing not at random, meaning imputation would not provide less biased results. It was suspected that lower values were associated with more advanced disease, and that blood tests were deferred in the sickest patients until they could consent to CD4 testing, meaning patients with the lowest values may have been less likely to have a baseline test. A sensitivity analysis in which all patients lost to follow-up were assumed to be either alive or dead did not alter the findings of either the 2 or 10-week model.

^a Only variables included in the multivariable model are shown. Adjusted for treatment, CD4 count, age, mental status, weight and fungal burden (see footnote b). Number included in final model = 413.

^b Peripheral white cell count, anemia, and raised CSF opening pressure were significantly associated with 10-week mortality after adjustment, but not included in the final model to prevent missing observations, markedly limiting the size of the model. Inclusion in a model considering only the patients with complete data (n = 391 for hemoglobin, n = 343 for peripheral white count, and n = 374 for raised CSF opening pressure) did not alter the magnitude or significance of the associations seen in the full model.

^c Test for trend.

^d QCC is shown in the univariable analysis as a categorical variable for ease of interpretation, but was included in the multivariable model as a continuous variable to give a better fit.