PI Governance SOP

1. Purpose and Scope

This Standard Operating Procedure governs the use of the ATLAS platform (Adherence Tracking and Longitudinal Assessment System) for multi-site medication adherence data collection in research studies. It covers workspace setup, role assignments, data collection procedures, quality assurance, and export protocols.

This document is intended to accompany the study's primary IRB protocol as a technical appendix describing the digital data collection platform. It does not replace the protocol itself.

Applicable study types

• Cross-sectional adherence surveys using the MMAS-8
• Longitudinal adherence tracking with repeat assessments
• Multi-site comparative studies with institutional sub-cohorts
• Quality improvement projects using adherence data for clinical decision support
• PEACS validation studies using the Predictive Emergence framework

2. Instrument Overview and Licensing

MAP (Multidimensional Adherence Parameters)

MAP (Multidimensional Adherence Parameters) is the 8-item, cross-sectional adherence instrument administered through ATLAS for institutional data collection. It provides a validated snapshot of medication adherence behaviour and is the intellectual property of Adherence Inc. (© TX 8-632-533). Use on ATLAS is governed by a licensing agreement between the study institution and Adherence Inc.

Every workspace key issued through ATLAS includes a Letter of Permission specific to the named PI, institution, and study title. This letter must be retained in the study's regulatory binder and submitted with ethics applications.

MAP PE Domain Analysis (TPE layer)

Available on the PI tier and above, the PE Domain Analysis applies the Theory of Predictive Emergence (TPE) to each patient's MAP response profile. It computes three behavioural domain scores from the Q1–Q8 responses already collected, then derives a composite PE score: PE = (A × E × C)^1/3.

• Architecture (A) — mean of Q2, Q3, Q6 — belief and motivation layer
• Execution (E) — mean of Q1, Q4, Q5, Q8 — habitual reliability layer
• Context (C) — Q7 — regimen burden and friction layer

PE domain values are stored on each assessment record and exported as mmas_pe, mmas_a, mmas_e, mmas_c columns in the MAP CSV. They are computed from the same Q responses already required for the standard instrument — no additional patient-facing questions are involved. PE is computable retroactively on all historical records in the cohort. The PI dashboard includes a PE Domain Analysis panel showing cohort-level domain averages and a lowest-PE patient table for intervention prioritisation.

PEACS v2.0

The Predictive Emergence Assessment for Clinical Services (PEACS v2.0) is a three-dimension instrument developed by Philip Morisky, Founder of Adherence Inc., based on the Theory of Predictive Emergence (DOI: 10.5281/zenodo.18209699). It is available on the PI and Institution tiers. PEACS data is collected at the same session as the MMAS-8 and stored in a separate data node.

Instrument validation status

The MMAS-8 has been validated in over 3,000 published studies across 100+ countries. MAP (Multidimensional Adherence Parameters) administered through ATLAS uses the original validated wording. The ATLAS platform itself has been used in institutional research at the University of Nicosia, Democritus University of Thrace, NUST Oman, and multiple clinical sites in Italy (SIMAT partnership).

3. Site Setup and Workspace Key Assignment

PI workspace (master key)

The PI receives a PI-tier workspace key (format: PI-[INST]-XXXX-2026) upon purchase at keys.adherence.cc. This key provides access to the full ATLAS dataset including all sub-site data, the multi-site dashboard, bulk upload, IRB-grade export, and PE Domain Analysis.

Site coordinator workspaces

Each data collection site should have its own workspace key (typically Researcher tier). This allows:

• Site-specific data isolation — each coordinator sees only their site's patients
• PI-level oversight — the PI's key sees all sites in aggregate
• Audit trail — each submission is tagged with the submitting workspace key

Sub-site keys are issued as Researcher tier and linked to your PI key as the Parent Institution Key — their data flows into your multi-site dashboard automatically. Contact info@adherence.cc to provision sub-site keys, providing: site name, coordinator name, and coordinator email. If your study is operating under an Institution account, sub-site Researcher seats are included in the institutional plan (5 seats in base, additional at $39/seat/month) — no separate subscription required per coordinator.

Setting up a new site

4. Role Hierarchy and Access Controls

5. Data Flow and Storage

Collection path

1. Patient opens ATLAS (directly or via shared link) on a device provided by the researcher
2. Patient reads and accepts digital consent
3. Patient completes MMAS-8 (and PEACS if included in protocol)
4. Submission is transmitted over TLS 1.3 to Google Firebase Realtime Database (Iowa, USA)
5. Record is tagged with the researcher's workspace key as institution_code
6. Record is immediately visible in the researcher's dashboard
7. If score is below threshold, Sentinel alert is written to Firebase alerts/{workspace} node

Data isolation

Each workspace key creates an isolated data namespace. A coordinator with key RES-SITE1-XXXX-2026 can only see records tagged with that key. The PI with key PI-INST-XXXX-2026 can see records tagged with either the PI key directly or any child workspace registered under it.

Data elements collected

21 CFR Part 11 — Electronic Records & Electronic Signatures

When ATLAS records are submitted to or used in support of FDA-regulated studies (IND, NDA, 510(k)) or GCP-compliant clinical research, the platform operates under FDA 21 CFR Part 11. The following controls are active in the platform:

For studies requiring FDA submission, contact info@adherence.cc to request the system validation documentation package (IQ/OQ/PQ) for inclusion in your regulatory binder.

6. Data Collection Procedures

Starting a patient session

Every assessment begins with the Start New Patient Session workflow. The session modal requires the researcher to select the instrument(s) before the patient touches the screen, ensuring the correct assessment pathway is loaded and a unique Patient ID is generated prior to consent.

Instrument selection in the session modal

Clicking Continue to Informed Consent → generates an auto-assigned Patient ID (e.g. PT-A3F7). The researcher must record this ID in the CRF immediately — it is the link between the ATLAS record and all other study data. This ID cannot be changed after the session begins.

Standard procedure — researcher-administered

1. Researcher logs into ATLAS with their workspace key
2. Researcher clicks Start New Patient Session and selects the appropriate instrument per protocol
3. Researcher clicks Continue to Informed Consent → and notes the auto-generated Patient ID in their CRF
4. Researcher selects the patient's preferred language on the SDoH screen (before questions begin)
5. Researcher either: (a) hands device to patient for self-completion, or (b) reads questions aloud and records answers
6. Patient submits assessment; session closes automatically
7. Researcher reviews result on dashboard and documents any Sentinel flags

Procedure for patients using their own devices (remote collection)

1. Researcher generates a direct assessment link: atlas.adherence.cc?key=[WORKSPACE-KEY]&pid=[PATIENT-ID]
2. Link is shared with patient via email, SMS, or patient portal
3. Patient completes assessment independently
4. Result appears in researcher's dashboard under the specified Patient ID

Deviations from protocol

Document any departure from this SOP in your study's deviation log, including: assessments completed without Patient ID assignment, technical failures requiring manual re-entry, or patients who completed assessment in incorrect language. These should be noted but do not invalidate the data.

6a. Clinical Study Module — Research Operations

The Clinical Study Module is accessed via the Study tab in the Institution dashboard. It provides a unified research operations view with the following components:

Study Configuration

Before collecting data, complete the Study Config (gear icon in the Study tab). Required fields: Study Name/Title, Principal Investigator, Institution, IRB Number. Optional but recommended: ClinicalTrials.gov ID, Protocol Version, Sponsor, enrollment target and visit schedule.

Site Activation Tracker

A world map displays all registered study sites colour-coded by status (Pending / Active / Closed). Add sites via the Add Site button. Site coordinates and status are stored locally and reflected on the map in real time.

Enrollment and LTFU

The module tracks enrolled patients against target, flags patients who are Lost to Follow-Up (LTFU) based on their expected visit window, and logs protocol deviations.

IRB Submission Assistant

Opens a 5-tab auto-populated IRB document template drawing from your Study Config. Tabs: Protocol Summary, Informed Consent, Risks & Benefits, Privacy & Data, Timeline & Sites. Each section is editable and the complete document exports as a structured .txt file suitable for submission or paste into your IRB system.

6b. Bulk Upload v2 Template — Study Provenance

Bulk upload allows researchers to import patient assessment records from a structured CSV file directly into the ATLAS platform. The current template version includes study-level metadata for regulatory traceability.

7. Quality Assurance

Weekly data review

The PI (or designated data manager) should review the dashboard weekly during active collection to:

• Confirm expected submission volumes across sites
• Review Sentinel alerts for clinically urgent cases
• Check for duplicate Patient IDs (same patient submitted twice)
• Verify geographic distribution matches enrollment plan

Identifying and handling duplicates

The Active Patients table can be searched by Patient ID. If the same Patient ID appears more than once with the same timestamp, this may be a duplicate. Contact info@adherence.cc to review and remove confirmed duplicates from your dataset.

Data completeness monitoring

The export CSV includes a timestamp column. Use this to verify all sites are submitting data within expected windows. Gaps may indicate technical issues, coordinator absences, or site-level protocol deviations that require follow-up.

8. PI Research Dashboard

The PI Research Dashboard is a dedicated panel at the top of your researcher view, visible only to PI-tier workspaces. It loads automatically when you sign in and consolidates the key governance metrics for your study in one place.

Enrollment velocity and target

The four KPI cells at the top of the panel show: total enrolled participants, current weekly enrollment rate (calculated over the last 8 weeks), projected completion date (based on current rate toward your target), and the data retention deadline (computed from study end date plus the mandated retention period for your funding source — 7 years for NIH/NSF/DoD/PCORI, 10 years for industry-sponsored, 5 years for all others).

Set your enrollment target using the target input field. The progress bar fills as you approach your goal. Projected completion updates in real time as new records arrive.

Cross-site PE Domain heatmap

Below the KPI strip, the PE heatmap shows one row per collection site (including your own PI workspace). Each cell displays the mean Architecture, Execution, and Context domain score for that site, colour-coded from red (constraint present) to green (domain healthy). This makes cross-site adherence barrier patterns immediately visible without exporting data.

Sites with fewer than 3 records are greyed out — insufficient data for meaningful domain averages. The heatmap refreshes each time the dashboard loads.

Dataset lock and frozen snapshots

The Dataset Lock section allows you to create an immutable copy of all assessment records at a specific point in time. Use this for:

• IRB interim analysis — lock the dataset at the analysis cutoff date before any interim review
• Collaborator data handoff — share the snapshot CSV rather than live access to your workspace
• Publication datasets — lock the exact records used in the submitted manuscript
• Study closeout — create a final locked snapshot before cancelling the subscription

Click ⊘ Lock Current Dataset to create a snapshot. The platform confirms the record count before locking. Each snapshot is permanently stored and listed with its creation date, MMAS count, and PEACS count. Use the ↓ CSV button next to any snapshot to download that frozen dataset.

Workspace audit log

The Audit Log displays the last 50 timestamped events for your workspace, newest first. Events recorded include: key issuance, key verification (each sign-in), data exports, dataset lock operations, site provisioning, electronic signatures, and failed login attempts. This log is suitable for inclusion in a regulatory binder as evidence of controlled access and data handling.

The audit log is append-only — entries cannot be deleted or edited. It persists for the lifetime of your subscription. All clinical record write operations (create, update, delete) generate a CFR-11 audit entry with a SHA-256 payload hash, satisfying FDA 21 CFR Part 11 §11.10(d)(e) requirements for tamper-evident, computer-generated, time-stamped records.

9. IRB-Grade Data Export and Dataset Lock

The PI tier includes IRB-grade export headers — additional metadata columns added to the CSV that are not present in the standard Researcher export.

Additional IRB export columns

Blinded export

The ⊘ Blinded Export button (in the MMAS-8 panel footer, visible to PI accounts) generates a CSV in which all patient IDs are replaced with deterministic hash codes. The same patient will always receive the same hash code within your workspace, so longitudinal records remain linkable — but the original patient numbers are not recoverable from the export. Use this format when sharing datasets with collaborators who do not need to know your patient numbering scheme.

Export procedure for study closeout

10. Escalation Matrix

11. IRB Submission Language

Methods section — data collection platform

Ethical considerations section

Data management plan section

FDA-regulated studies — 21 CFR Part 11 language

Use this section for IND, NDA, 510(k), or GCP-compliant studies where the FDA 21 CFR Part 11 electronic records requirement applies.

12. Required Citations and Acknowledgements

MAP — required citation in all publications

Required acknowledgement footnote

ATLAS platform citation

PEACS — if used in study

12a. Publication Rights and Licensing

What your Letter of Permission covers

Your Letter of Permission (LOP) is issued at workspace key activation and records your subscription activation date. It covers data collected during your active subscription period only. It does not reach backward. If your subscription lapses at any point, the LOP expires for that gap period — data collected after reactivation is covered from that new activation date forward.

Multi-year and grant-funded studies

For PI-tier studies funded by grants, contracts, or institutional budgets, a standard subscription covers your data collection period while the subscription is active. If your study spans multiple years and you maintain continuous subscription coverage, your LOP remains valid throughout. If your study began before you subscribed to ATLAS, only the post-activation portion of your data is covered under the subscription LOP.

If you need publication rights covering a study that predates your ATLAS subscription — for example, you conducted a funded study before the platform was available and are now seeking to publish — contact info@adherence.cc with your study title, funding source, IRB number, and data-collection dates. Publication rights for funded studies are priced by study duration, not a flat rate.

Two publication rights paths

What happens if I cancel my subscription before publishing?

If your subscription lapses before your manuscript is submitted, your Letter of Permission expires. To restore publication coverage you have two options: reactivate your subscription (LOP reissued immediately, covers data from reactivation forward), or contact info@adherence.cc to discuss duration-based publication rights for the completed study period.

13. SOP Amendments

This SOP should be updated when any of the following occur:

• A new version of the ATLAS platform introduces materially different data collection procedures
• The MAP license terms are amended by Adherence Inc.
• IRB conditions require modification to collection procedures
• A new data collection site is added to the study

All amendments should be dated, versioned, and re-signed. Previous versions should be retained in the study regulatory binder. Notify info@adherence.cc of any material amendments to ensure platform support is aligned with updated procedures.