Test for a trend of bacterial resistance against antibiotics over ten years period

[Abdullah Algarni]

Dear statalist,

Unfortunately, I posted this question one week on cross-validated and have yet to get an answer (see here please: https://stats.stackexchange.com/q/603421/362674?sem=2).
We are analyzing a dataset of Acinetobacter bacteria incidence and their resistance pattern against antibiotics over ten years (2013-2022). Every year, there are multiple resistant isolates, but that has a unique ID no. (i.e., stata_id in our dataset).

Study hypotheses:

• Null hypothesis: The trend of Acinetobacter resistance against antibiotics is stable (constant) over the study period.
• Alternative hypothesis: The trend of Acinetobacter resistance against antibiotics changed (decreased/increased) over the study period.

We are using Stata 17 SE.
The structure of our dataset is as follows:

* Example generated by -dataex-. For more info, type help dataex
clear
input float stata_id byte year float pc_cra
   1  1 75.82089
1898  7 87.38171
2521  9 89.03509
1672  6 94.06393
 100  1 75.82089
2032  7 87.38171
2519  9 89.03509
2806 10  75.6654
2700 10  75.6654
2353  9 89.03509
 193  1 75.82089
 689  2 85.35912
1200  4 84.98293
 258  1 75.82089
1495  5 88.30189
  29  1 75.82089
1117  4 84.98293
1905  7 87.38171
 836  3 91.36213
 529  2 85.35912
end
format %ty year
label values year Year
label def Year 1 "2013", modify
label def Year 2 "2014", modify
label def Year 3 "2015", modify
label def Year 4 "2016", modify
label def Year 5 "2017", modify
label def Year 6 "2018", modify
label def Year 7 "2019", modify
label def Year 9 "2021", modify
label def Year 10 "2022", modify

Description of variables:

. fre cra 

cra -- Carbapenems resistant?
-----------------------------------------------------------
              |      Freq.    Percent      Valid       Cum.
--------------+--------------------------------------------
Valid   0 No  |        394      13.93      13.93      13.93
        1 Yes |       2434      86.07      86.07     100.00
        Total |       2828     100.00     100.00           
-----------------------------------------------------------

​​​​​​​. fre pc_cra 

pc_cra -- Percent of CRAB isolates/year
--------------------------------------------------------------
                 |      Freq.    Percent      Valid       Cum.
-----------------+--------------------------------------------
Valid   75.6654  |        263       9.30       9.30       9.30
        75.82089 |        335      11.85      11.85      21.15
        84.98293 |        293      10.36      10.36      31.51
        85.35912 |        362      12.80      12.80      44.31
        87.38171 |        317      11.21      11.21      55.52
        88.30189 |        265       9.37       9.37      64.89
        89.03509 |        228       8.06       8.06      72.95
        91.36213 |        301      10.64      10.64      83.59
        93.06123 |        245       8.66       8.66      92.26
        94.06393 |        219       7.74       7.74     100.00
        Total    |       2828     100.00     100.00           
--------------------------------------------------------------

​​​​​​​. fre year

year -- Year of culture
-------------------------------------------------------------
                |      Freq.    Percent      Valid       Cum.
----------------+--------------------------------------------
Valid   1  2013 |        335      11.85      11.85      11.85
        2  2014 |        362      12.80      12.80      24.65
        3  2015 |        301      10.64      10.64      35.29
        4  2016 |        293      10.36      10.36      45.65
        5  2017 |        265       9.37       9.37      55.02
        6  2018 |        219       7.74       7.74      62.77
        7  2019 |        317      11.21      11.21      73.97
        8  2020 |        245       8.66       8.66      82.64
        9  2021 |        228       8.06       8.06      90.70
        10 2022 |        263       9.30       9.30     100.00
        Total   |       2828     100.00     100.00           
-------------------------------------------------------------

We have the following questions:

1. Is our data time-series data, panel data, or something else?
2. What is (are) the appropriate statistical test(s) to use for testing the trends over the years?

Thank you

[Carlo Lazzaro]

Abdullah:
I find your description and your data inconsistent, because:
1) you state that each year the analysis is performed, the resistant isolates have the very same id. Therefore, it seems you're dealing with a panel dataset. But looking at your data, no -id_stata- ia actually repeated (hence, as per your example, this would seem a cross-sectional study);
2) the appropriate statistical tools are conditional on clarfying point 1).

[Abdullah Algarni]

Abdullah:
I find your description and your data inconsistent, because:
1) you state that each year the analysis is performed, the resistant isolates have the very same id. Therefore, it seems you're dealing with a panel dataset. But looking at your data, no -id_stata- ia actually repeated (hence, as per your example, this would seem a cross-sectional study);
2) the appropriate statistical tools are conditional on clarfying point 1).

Thank you Carlo,
I said "Every year, there are multiple resistant isolates, but that has a unique ID no. (i.e., stata_id in our dataset)."
So, there is no repeated stata_id.

Thank you

[Carlo Lazzaro]

Abdullah:
thansk for clarifying and apologise for my previous misreading.
So yiou have a repeated (in years) cross-sectiona study.
Perhaps you may want to consider a logisic regression with the dependent variable being -ab_resistance- (yes=1; no=0); -i.antibiotics- and -c.year##c.year- as predictors:

logit ab_resistance i.antibiotics c.year##c.year

[Abdullah Algarni]

Thank you so much Carlo