DCLP5

This page summarizes our insights about the clinical study data of the DCLP5 study in efforts to understand how to handle bolus, basal, and cgm data, list assumptions that were made, and provide open questions.

The full analysis of this dataset is provided in: notebooks/understand-dclp5-dataset.ipynb

Study Overview

• Study Name: The International Diabetes Closed Loop (iDCL) trial: Clinical Acceptance of the Artificial Pancreas in Pediatrics - A Pivotal Study of t:slim X2 with Control-IQ Technology
• Study Background: The objective of the study was to assess efficacy and safety of a closed loop control (CLC) system (t:slim X2 with Control-IQ Technology) in a randomized controlled trial with partial crossover.
• Study Devices: t:slim X2 with Control-IQ and Dexcom G6 system
• Study Population: Children aged 6 - 13 years old with type 1 diabetes
• Total Data: There are roughly 19,200 days of data from 100 participants

Data

The study data folder is named DCLP5_Dataset_2022-01-20-5e0f3b16-c890-4ace-9e3b-531f3687cf53

While there is a data glossary (DataGlossary_Dits.rtf) file, the file names within the glossary do not match the file names of the data. However, the column names do match and those definitions are listed below.

These are csv files ("|" separator) and host many columns related to the Tandem pump events and the Dexcom cgm. The glossary provides information about each column. Each file contains a limited amount of columns compared to the FLAIR data.

Bolus

• DCLP5TandemBolus_Completed_Combined_b.txt: List of pupm data downloaded (bolus data only)

Column	Description	Comment
PtID	Patient ID
DataDtTm	Date-time of delivered insulin
BolusAmount	The amount of insulin actually delivered for this bolus	In Glossary this is called InsulinDelivered
BolusType	The bolus insulin delivery type [Standard, Extended]
DataDtTm_adjusted	adjusted Date-time	Not mentioned in glossary

Basal

• DCLP5TandemBASALRATECHG_b.txt: List of pump data dowloaded (basal data only)

Column	Description	Comment
PtID	Patient ID
DataDtTm	Date-time of basal rate change
CommandedBasalRate	Basal Rate (U/h) - The active basal insulin delivery rate in units per hour
DataDtTm_adjusted	adjusted Date-time	Not mentioned in glossary

CGM

There are three CGM files mentioned in the glossary:

• DCLP5DexcomClarityCGM (actual name DexcomClarityCGM.txt): CGM readings from Dexcom Clarity (G5) files -- Final table with duplicates removed
• DCLP5OtherCGM (actual name: OtherCGM.txt): List of CGM readings from CGM file types for which loading processes have not been set up (e.g. Tandem CGM hourly report, or CGM tab from Diasend file saved as csv, etc)
• Pump_CGMGlucoseValue (actual name: DCLP5TandemCGMDATAGXB_b): CGM glucose value recorded on the pump

• TandemCGMDATAGXB.txt (this was not mentioned in the data glossary)

• DCLP5TandemCGMDATAGXB_b.txt: List of cgm data dowloaded 

Column	Description	Comment
PtID	Patient ID
DataDtTm	Date-time
CGMValue	40-400 mg/dL. 40-400 are legitimate glucose values. 0 is used when the reading is high or low. Use of this field depends on the glucoseValueStatus field.	Same as in DCLP3
HighLowIndicator	0 = CGMValue contains the glucose reading. 1 = The glucose reading is high (CGMValue set to 0). 2 = The glucose reading is low (CGMValue set to 0).
DataDtTm_adjusted	Adjusted Date-time

Observations

• Very confusing differences between Glossary and actual filenames and column names.
• e.g. there are 4 CGM Files
• The glossary file names don't match the actual file names
• The glossary column names are named differently; some are missing
• Overall structure as in DCLP3, but different file names

DCLP5 vs. 3 (Datafiles)

This is an overview of file names and glossary discreptancies within and between DCLP3 and 5. Test that ~~crossed through~~ are names were incorrect.

	DCLP3	DCLP5
Clarity	DexcomClarityCGM M readings from Dexcom Clarity files	DCLP5DexcomClarityCGM CGM readings from Dexcom Clarity (G5) files -- Final table with duplicates removed
Other	OtherCGM List of CGM readings from other miscellaneous sources cgm aggregation of CGM data from multiple sources within the study that was used for the analyses presented in the primary manuscript DiasendCGM Readings from CGM tab of Diasend file	DCLP5OtherCGM List of CGM readings from CGM file types for which loading processes have not been set up (e.g. Tandem CGM hourly report, or CGM tab from Diasend file saved as csv, etc)
Pump	Pump_CGMGlucoseValue One record per CGM glucose value recorded on the pump	~~Pump_CGMGlucoseValue~~ DCLP5TandemCGMDATAGXB_b.txt CGM glucose value recorded on the pump
Conclusion	We used this one: Pump_CGMGlucoseValue.txt	None of the files exists in reality. These are the files in the data table folder: DCLP5TandemCGMDATAGXB_b.txt (we used this one) TandemCGMDATAGXB.txt DexcomClarityCGM.txt OtherCGM.txt

Basal:

	DCLP3	DCLP5
File	Pump_BasalRateChange.txt	~~Pump_BasalRateChange.txt~~ DCLP5TandemBASALRATECHG_b.txt
Column	~~BasalRate~~ CommandedBasalRate	CommandedBasalRate
Adj. Date	DataDtTm_adjusted	DataDtTm_adjusted (Not Mentioned)

Bolus:

	DCLP3	DCLP5
File	Pump_BolusDelivered.txt	~~Pump_BolusDelivered~~ DCLP5TandemBolus_Completed_Combined_b.txt
Insulin	BolusType	~~InsulinDelivered~~ BolusType
Type	~~InsulinDelivered~~ BolusAmount	BolusAmount
Date	DataDtTm	DataDtTm
Adj. Date	DataDtTm_adjusted	DataDtTm_adjusted (not mentioned)

Data Integrity

Duplicates & Missing Data

• No NaN values, except the adjusted datetimes (expected)
• There are no complete row duplicates (e.g. also considering record id)

When only checking for PtID, DataDtTm, and DataDtTm_adjusted:

• 4290 basal rates reported at the same time (but most of them are equal)
• 3262 cgm duplicates (but most share the same value)
• 124 bolus duplicates (but all have different bolus types) (extended, normal) --> no real duplicates

We deal with duplicates as follows:

• Basal: Taking the maximum value (we see some zero values)
• CGM: Taking the first value
• Bolus, we handle later in the Extended Bolus section

Incomplete Patients

There is no need to filter patient data as all patients have complete data.

Datetimes

In Flair and PEDAP datetime strings were reported in inconsistent time formats: time components were missing at midnights which required special handling.

We see that:

• there are no missing datetimes
• As in Flair, datetimes sometimes miss the time component, even if it is only a few instances
• The date formats are different from DCLP3 and PEDAP
• DCLP3: %Y-%m-%d %H:%M:%S
• DCLP5/PEDAP: %-m/%-d/%Y %I:%M:%S %p

We therefore again check if the missing time components align with midnight by checking a few examples: What we see is that the 5th value (which is the one with the missing time component) snuggly fits the series. Therefore, we can safely assume it is in fact the midnight value. The drops to zero are just because because they are out of range.

For performance purposes we use the parse_flair_dates() function to parse the datetimes.

Datetime Adjustments

Again, we test if using the adjusted datetimes, if available, is appropriate. In the other datasets this was always true but we checked again to make sure. We see see that using the adjusted datetimes moves the cgm data in the right spot (mostly) and reduces the variation in time differences. DCLP5 cgm with missing time component treated as midnight fits into overall data

Timestamps

Time Localization

The DCLP3 data glossary makes no mention of UTC, timezone ,or zone. All datetime variables are described as local times. We want to verify by checking the distribution of mean CGM, basal rates, and bolus doses to verify we see postprandial peaks in the morning, afternoon, and evening as well as more stable glucose during the night.

We see bolus peaks at around 7am, noon, and evening. Slightly delayed, we see postprandial glucose peaks as expected. Due to Control IQ, basal increases significantly after dinner then slowly decreases overnight. This all makes sense and we can safely assume that all datetimes are in local time.

CGM

CGM data

Replace Zero CGMs

From the data glossary we know that 0 cgm values are either below or above range.

0 =CGMValue contains the glucose reading
1 =The glucose reading is high~ CGMValue set to 0
2 =The glucose reading is low~ CGMValue set to 0

Special Values

- Identify and replace any special CGM values if needed.

Boluses

There are no bolus duplicates when including the BolusType column. The duplicates in boluses seem to be a result of Standard and Extended Bolus portions being reported at the same time. Is this true for all extedned boluses? Let's check.

There are 3169 extended boluses (~1.4% of the boluses). We apply the same logic as in DCLP3 for matching standard and extended boluses.

We found 3136 periods for 3169 extended boluses Mean (STD) [MEDIAN] of extended boluses: 63.72 (49.83), [55.47] minutes

In comparison DCLP3: Mean (STD) [MEDIAN] of extended boluses: 59.02 (46.28), [55.42] minutes

As in DCLP3, We won't be able to clearly separate dual wave boluses (standard followed by an extended part) from orphan extended boluses (without standard part): Orphan Extended boluses will be incorretly matched to the previus standard bolus.However, this will happen only to a small percentage (In PEDAP only ~2% (170 out of 6460)) extended boluses did not have a standard bolus part. The T-Slim X2 allows durations <=8 hours, therefore pairs matched with durations >8 hours are definitely wrong.

Moving forward, we will 1. assume that extended boluses are reported upon completion (as in PEDAP) 2. Estimate delivery_duration: - Standard boluses: set to 0 - Extended boluses: time difference to their matched standard boluses - Durations >8h (those are definitely matched incorrectly) will be set to 55 minutes (median value)

Basal Rates

In PEDAP, we confirmed that basal rate change events are reported whenever basal delivery changes, not only if the new basal rate deviates from standard basal rate. Here, we apply the same logic and assume all basal rate changes are present.

Suspends

There are no suspend events which we know of that need to be considered.

Temporary Basals

There are no temporary basal rates that we know of that would need to be considered.