# [R] Hospital ANOVA/ANCOVA problem

edamamebuffet alanhfung at gmail.com
Wed Nov 24 21:18:13 CET 2010

```Hi everyone,

I've only been using R for a week or so now, and am now required to work
with some pretty mixed data at my hospital.

Basically, we're looking to see if hypertension status (nominal) affects
cardiac function (multiple & continuous). In order to do this, we've
collected multiple echocardiograph data reports on each patient.

1. Each patient is coded either as 0 or 1 for hypertension (HT).
2. Each patient then has multiple echo reports (numbering anywhere from 1 to
20) with variables (A, B, C, D, E, F, etc) for each report
3. I want to see if A, B, C, D are individually correlated to hypertension
status.

The problem is that we must also take into account how old the echo report
is since the heart takes time to remodel/change.

Independent variables: hypertension status (HT - nominal) and echo time
since transplant (ECHO_DAYS - continuous)
Dependent variables: echo report parameters, some of which may be related.
(A, B, C, D, E, F, etc - continuous)

The dataset looks like this:

ID    HT    ECHO_DAYS    A    B    C    D
1    0    23    11    25    19    45
1    0    50    77    37    43    85
1    0    70    NA    8    NA    95
1    0    100    89    16    70    82
1    0    134    44    54    76    58
2    1    12    14    93    86    77
2    1    23    39    53    87    38
5    1    30    NA    NA    NA    53
5    1    39    89    15    19    95
5    1    40    35    9    63    81
5    1    41    47    3    94    NA
7    0    48    97    68    66    NA
7    0    32    36    86    29    69
9    1    34    53    19    34    86
9    1    1174    86    62    40    73
9    1    1880    43    84    35    88
9    1    370    17    19    54    6

Currently I'm using:
aov.a2 =aov(A~(HT+ECHO_DAYS)+Error(ID/(HT+ECHO_DAYS)))

1. Can anyone verify that this usage is correct?
2. If not, what is the best way to calculate and graph the ANOVA/ANCOVA I'm
looking for.
3. Please excuse my ignorance. Again, statistics is not my forte...

Thank you!
--
View this message in context: http://r.789695.n4.nabble.com/Hospital-ANOVA-ANCOVA-problem-tp3058020p3058020.html
Sent from the R help mailing list archive at Nabble.com.

```