Post-event Pain algorithm

Pain is a personal, multidimensional experience in which genetic biomarkers has a main role in determining pain sensitivity, perception and tolerance. Pain is a major concern for surgical patients and post-operative pain management still present a major challenge both in inpatient or outpatient settings. Apart from genetic factors, there are many other variables that may affect pain perception for example, pretreated patients may require less post-surgical medications, and they may recover more quickly. Although there are no gold standard, numeric rating scale (NRS 0–10) are regularly used in postoperative pain treatment. The estimated tolerable postoperative pain before operation is median of NRS 4.0 (0–10). Patients who would need more analgesics reported significantly higher average pain since surgery median NRS 5.0 compared with those without this request NRS 3.0 (1).

Phenotype ID: 
Do Not List on the Collaboration Phenotypes List
Contact Author: 
Date Created: 
Thursday, September 1, 2016
Network Associations: 
Owner Phenotyping Groups: 
View Phenotyping Groups: 

Suggested Citation

CCHMC. Post-event Pain algorithm. PheKB; 2016 Available from:


Sickle cell questions:

For case one sickle cell, do we just report on the first hospitalization with pain scores?
Also, is it correct that we only report medications on case ones since only those have NRS scores?

Sickle cell and surgery questions:

The encounter meds DD age is age at date of CPT code. Should that be age at NRS score or age at medication? 
Report only pain meds or all meds?

All medications at time of hosptitalization need to be recorded.

If the sickle cell patient has multiple hospitalizations, please record both medication and NRS (if available). 

For age, the intent is to be the age at the hospitalization or procedure. So the first encounter date/age is appropriate, if that's what is available. Otherwise age at the NRS score or medication is acceptable, provided it was in the same encounter with the CPT/procedure.

Thanks Todd, we will report on multiple hospitalizations for sickle cell cases.  However the DD for NRS scores has a column for minutes from first NRS score but no way to distinguish one hospitalization from another. Also the current meds DD does distinguish hospitalizations but we'll have duplicate rows if the same med/dose/route is given more than once during a hospitalization (we may also have cases where the same meds are given more than once on the same day).

We will update the NRS DD to indicate age at time of NRS score

We will update the medication to indicate age at time ordered

We can do the time-analysis necessary to off set the days after encounter (expecially given the LOS days in the NRS DD).

How should we report Stroke, infarction, acute chest pain, Gallstone (cholecystitis), Dactylitis?  Can you provide codes for these?

Do the number of x's for the ICD10 codes indicate a required digit, for example does D57.21xx require 4 digits to the left of the decimal point?  Also would G82.x require 1 but two are OK?

If we have SC candidates for cases or controls who are not clearly indicated as African American or who are self-reported as Asian, Caucasian or Other, should they be included or excluded?

The algorithm description file has been updated to use * as standard for all available codes/digits after that position.


Do we apply any timeline criteria to the NRS data in relation to the event (surgery or sickle cell crisis) or do we report all NRS records we have on the patient? My understanding is that we would report all events along with medication data at the time of the event, along with all NRS records patient has (which may or may not coincide with the event dates) .



Thanks for the question. The NRS scores need to be during the encounter with either the surgery or the crisis admission (sickle cell).

Submitted by Xinnan Niu on

The document states for case,  any individuals above age 6 years old who undergone one-time surgery for scoliosis or pectus excavatum, How about an individuals above age 6 years old but undergone more than one-time surgeries, how can we report ? 

Submitted by Xinnan Niu on

There are two tables, Table 2 - Exclusion list and Table 3 Covaritat Diagnoses, over the section "Sicele Cell Pain" but they were not menotioned in the context of that docuemnt.  What are the codes listed in Table 2 and 3 from Sicle Cell Pain used for ? 

Table 2 in Sickle Cell is Case Exclusion.

Table 3 is Covariates used in the Sickle Cell Covariate Dx data dictionary.

Submitted by Xinnan Niu on

For table 3, Table 3—Covariate Diagnoses, I guess it was desigined for the implemented binary values defined in SickleCell_CovariateDx_v3.csv, correct ? If wrong, please correct me.  However, I still am having difficulty to figure out the application/purpose of Table 2-Exclusion list. What is used to for ? Would you please clarify it, Thanks,

Submitted by Xinnan Niu on

Thanks Todd, and would you explain what are those codes used for exclusion defined in table 2  for section,Sicele Cell Pain ?

Submitted by Xinnan Niu on

The field, GWAS_SEQ, defined in Pain_DataDicitionary_demographics_v2-2.csv has 0=GWAS;1=SEQ;2=GWAS_AND_SEQ.

For SEQ, it means patients who got one of eMEGREseq, PGX PGRNseq, WGS (whole genome seq), WES (whole exome seq) are entitle as value "1", correct ?

Submitted by Xinnan Niu on

How is the age <6 years is defined in tabl-2 exclusion list ? Is it the earliest age when a patient got one of the criteria of defined codes/medication ? Need to clairify  and thanks !

Yes, earliest age of inclusion criteria. 

Submitted by Xinnan Niu on

Question : what is the vaule we should to give for LOS ?  if a subject was coded with procedure codes but without the encouter infomation to calculate LOS, should we give the missing value '.' if yes, the DD sheet was not defined this.


"." is appropriate for missing length of stay information. I'll update the data dictionary. 


Submitted by Xinnan Niu on

Question #1. The algorithm used to define CASE in the doucment (Post-event Pain algorithm V5.docx) only specified to use a list of ICD9 in Table 1a but not ICD10. Need to confirm that forget to update the context of words but only updated with ICD10 code in table 1a, right ? If yes, we need to add both ICD9 and 10 to  define the 3 categorized CASES.

Question #2. It is hard to get NRS scores from our EHR. Therefore, we will skip the case 1, which have both encounter information and NRS scroes for generating mean,max,SD, and range. By this way, the cohort of SCP from our site only contain case 2 and case 3, Is it OK ?

Question #3. Since we do not have informaton for NRS score, it will hard to fully implement the DD sheet, Pain_NRS_v4-2.csv but we can patially implement with a list of subject and rest of fileds with missing values a '.'. Is this OK ? if yes, I will do it but you might need to update that sheet with '.' for the rest of fields, which we are not able to provide.



#1 ICD9 OR ICD10 are sufficient. Both aren't necessary. I don't see ambiguity in the text of the document. 

#2,3 Sickle Cell is the only case that can be permitted without NRS scores. The other primary case for post pain procedure doesn't work without NRS scores. Please do not submit patients without meeting SCP criteria if you do not have NRS scores.



In table Table 3b and 4b, there only ICD-9 codes provided. Are we not using any ICD10 equivalents for these codes?



Submitted by Xinnan Niu on

Thanks Ozan ! I mean the ICD 10 codes listed in table 1a and 1b for Sickle Cell Pain but not Table 3b and 4b for Post event pain. There are icd10 codes listed in table 1a and 1b but the docoument of SCP only specified to use ICD9 but not ICD10 codes.


ICD9 PROC and ICD9 aren't the same.
Not all ICD-procedure codes has not been accurately converted yet. Please follow your own conversion policy considering these icd9 procedure codes.

I don't understand the question. If they are not in GWAS and not in the eMERGE SEQ, why are they being included? Are they in your eMERGE cohort without having had genetic information submited?

Submitted by Xinnan Niu on

Thanks Todd and yes they are in the emeger cohort but without genetic information submitted. 1). they might be dropped out from the cohort without updating database information. 2). the genetic information might be updated but the database was not updated as well. Let me check.  I will drop those patients from cohort if I confirm that they were truely elmimated from emeger cohort.

Submitted by Xinnan Niu on

Hi Todd, I got the updated emeger cohort data for gwas and sequnce. By searching the updated sheet, I found

1). There are 3 patients who are SCP cases and enrolled in emerge cohort but do not have information for GWAS and Sequencing.

2). There are 8 patients who are SCP cases and enorlled in emerge cohort but do not have information for GWAS and Sequencing.However, they did get genotyped on exomechip. What should we treat them ? they were not enrolled for GWAS and sequnce but got genotyped on exome chip and I do not think they are qualified as people's genome/patial genome got sequnced.

Can you clarify 'exomechip'? Where are these patient's genetic data? If they aren't available then you shouldn't include them.

Submitted by Xinnan Niu on

Todd, those patients got genotyped in terms of SNPs but not got sequenced, meaning  we get their gentic SNPs profile but not the genome sequnced data. Should we include them ? If not, I will eliminate them.


List of covariates in the word document for sickle cell phenotype differs from data fields in SickleCell_CovariateDx_v3.csv. How should we proceed?



Hi, can you tell us which ones you've found different?

Sickle cell covariates data dictionary file has the following fields; SUBJID, ICD9_Code, ICD10_Code, Age. 

Word document outlines following covariates for Sickle cell phenotype:  

1) History of conditions (Table 3)

2) Age
3) Gender
4) Race/ethnicity (primarily African-American)
5) Length of stay in hospital
6) Medication including route and dosage at the time of NRS records
7) # POD NRS (Number of Post-Operative Days of NRS scores in an encounter), max of all event encounters, if multiple.

We were not sure in which file to provide these data. Thanks for the help.