Comparative Temperature Analysis

Generated 2026-07-02 19:36 · Patient 1 (post-HSCT) vs Patient 2 (post-Stroke) · Data Jan 08, 2026 to Apr 14, 2026 · Day 30 post-treatment

P1 MEAN TEMP

In range

0.0C

n=9 days

P2 MEAN TEMP

In range

0.0C

n=449 days

P1 TEMP SD

Info

0.000C

IQR=0.000

P2 TEMP SD

Info

0.614C

IQR=0.000

ANOMALY DAYS

In range

1.0days

H=0, M=1

TRENDS

Temperature Trends

DISTRIBUTION

Distribution Comparison

Cross-patient comparison: Mann-Whitney U p=p=0.900, Cohen's d=-0.05 (negligible). Patient 2 runs warmer on average. Patient 2 shows more temperature variability.

RUX

Ruxolitinib Effect (Patient 1)

EARLY WARNING

Early Warning Model

Method: Spearman rank correlations at lags 0, 1, and 2 days. Lag 1 tests whether today's temperature predicts tomorrow's readiness or heart rate. A significant negative correlation at lag 1 (temp up, readiness down next day) suggests temperature as an early warning signal for declining health.

Patient	Target	Lag (days)	Spearman rho	p-value	n	Significant
P2	Readiness Score	0	-0.082	p=0.084	449	No
P2	Readiness Score	1	-0.060	p=0.205	448	No
P2	Readiness Score	2	-0.039	p=0.414	447	No
P2	Average HR	0	0.081	p=0.087	448	No
P2	Average HR	1	0.039	p=0.412	447	No
P2	Average HR	2	0.060	p=0.203	446	No
P3	Readiness Score	0	nan	N/A	22	No
P3	Readiness Score	1	nan	N/A	21	No
P3	Readiness Score	2	nan	N/A	20	No
P3	Average HR	0	nan	N/A	22	No
P3	Average HR	1	nan	N/A	21	No
P3	Average HR	2	nan	N/A	20	No

CALENDAR

Temperature Calendar Heatmap

EVENTS

Known Events Validation

Patient	Date	Event	Expected	Actual	Validated
P2	2022-05-04	+13C flight day	13.0	13.00 C	Yes
P1	2026-02-09	Acute event (Feb 9)	None	No data	No

Top Temperature Spikes (> +1.0 C)

Patient	Date	Temp Delta	Classification
P2	2022-05-04	+13.00 C	Fever

INTERPRETATION

Clinical Interpretation

Anomalies: 1 total anomaly days detected (Patient 1: 0, Patient 2: 1) using 2-SD threshold.

METHODOLOGY

Methodology

Data Source: Oura Ring sleep temperature deviation (temperature_delta) measures nightly skin temperature relative to the wearer's personal baseline, in degrees Celsius. Also incorporates readiness temperature_deviation.

Anomaly Detection: Days where |temperature_delta| exceeds 2 standard deviations from the patient's own mean are flagged as anomalies.

Classification: Fever (>+1.0 C), mild elevation (+0.5 to +1.0 C), normal (-0.5 to +0.5 C), mild low (-1.0 to -0.5 C), hypothermia (<-1.0 C).

Early Warning Model: Spearman rank correlations at lags 0, 1, and 2 days test whether temperature deviations predict changes in readiness score or heart rate 1-2 days later.

Treatment Analysis: Pre/post Ruxolitinib comparison uses Mann-Whitney U on |temperature_delta| and Levene's test for variance equality.

Cross-patient Comparison: Mann-Whitney U test with Cohen's d effect size and 95% bootstrap confidence intervals (10,000 iterations).