Incident Response & Disaster Recovery
Overviewโ
HIPAA requires documented procedures for security incident response (45 CFR ยง 164.308(a)(6)) and contingency planning (45 CFR ยง 164.308(a)(7)).
1. Roles and Responsibilitiesโ
| Role | Responsibility |
|---|---|
| HIPAA Privacy Officer | PHI use oversight, patient rights, breach notification |
| HIPAA Security Officer | Technical safeguards, incident response lead |
| Cloud Administrator | GCP infrastructure response |
| Legal Counsel | Breach notification compliance, regulatory reporting |
Fill in contact details before going live. Store in a location accessible without GCP (e.g., physical binder + encrypted offline document).
2. Incident Classificationโ
| Severity | Description | Response Time |
|---|---|---|
| P1 โ Critical | Active breach, ongoing data exfiltration | < 1 hour |
| P2 โ High | Confirmed PHI exposure, account compromise | 4 hours |
| P3 โ Medium | Suspected breach, anomalous activity | 24 hours |
| P4 โ Low | Policy violations, near-misses | 72 hours |
3. Incident Response Processโ
Detect โ Contain โ Analyze โ Eradicate โ Recover โ Document โ Notify
Step 1 โ Detectโ
# Query for anomalous DB access during investigation
gcloud logging read \
'protoPayload.authenticationInfo.principalEmail!="[email protected]"
AND resource.type="cloudsql_database"' \
--project=YOUR_PHI_PROJECT_ID --format=json
Step 2 โ Containโ
# Revoke a compromised service account
gcloud iam service-accounts disable \
COMPROMISED_SA@YOUR_PROJECT_ID.iam.gserviceaccount.com \
--project=YOUR_PHI_PROJECT_ID
# Remove a user's IAM access
gcloud projects remove-iam-policy-binding YOUR_PHI_PROJECT_ID \
--member="user:[email protected]" --role="roles/ROLE_NAME"
# Block an IP via Cloud Armor
gcloud compute security-policies rules create 900 \
--security-policy=phi-waf-policy \
--expression="inIpRange(origin.ip, 'ATTACKER_IP/32')" \
--action=deny-403 --project=YOUR_PHI_PROJECT_ID
Step 3 โ Analyzeโ
# Export audit logs for the incident time window
gcloud logging read \
"timestamp >= \"2026-02-20T00:00:00Z\" AND timestamp <= \"2026-02-20T23:59:59Z\"" \
--project=YOUR_PHI_PROJECT_ID --format=json > incident_logs.json
Step 5 โ Recoverโ
# List available backups
gcloud sql backups list --instance=phi-db-instance --project=YOUR_PHI_PROJECT_ID
# Restore from backup
gcloud sql instances restore-backup phi-db-instance \
--backup-id=BACKUP_ID \
--restore-instance=phi-db-restore \
--project=YOUR_PHI_PROJECT_ID
4. HIPAA Breach Notification Requirementsโ
Notification Timelineโ
| Notification | Recipient | Deadline |
|---|---|---|
| Internal escalation | Privacy Officer, Security Officer | Immediately |
| Affected individuals | Patients whose PHI was breached | Within 60 days |
| Secretary of HHS (< 500 affected) | HHS.gov portal | Within 60 days after year end |
| Secretary of HHS (โฅ 500 affected) | HHS.gov portal | Within 60 days of discovery |
| Media notice (โฅ 500 in a state) | Prominent media outlet | Within 60 days |
Required Notification Contentโ
- Brief description of the breach
- Types of PHI involved
- Steps individuals can take to protect themselves
- What you are doing to investigate, mitigate, and prevent recurrence
- Contact information
HHS Reporting Portal: https://ocrportal.hhs.gov/ocr/breach/wizard.jsf
5. Disaster Recovery Planโ
Recovery Objectivesโ
| Metric | Target |
|---|---|
| RTO (Recovery Time Objective) | 4 hours |
| RPO (Recovery Point Objective) | 1 hour |
Backup Architectureโ
Primary Region (us-central1)
โโโ Cloud SQL (PRIMARY)
โ โโโ Automated daily backup โ Cloud Storage
โ โโโ PITR: 7-day transaction log window
โ
โโโ Redis (STANDARD_HA)
โโโ Automatic failover to standby replica
DR Runbook โ Cloud SQL Failureโ
# 1. Check instance state
gcloud sql instances describe phi-db-instance \
--format="value(state)" --project=YOUR_PHI_PROJECT_ID
# 2. List backups
gcloud sql backups list --instance=phi-db-instance --project=YOUR_PHI_PROJECT_ID
# 3. Restore
gcloud sql instances restore-backup phi-db-instance \
--backup-id=LATEST_BACKUP_ID \
--restore-instance=phi-db-recovery --project=YOUR_PHI_PROJECT_ID
DR Test Scheduleโ
| Test Type | Frequency |
|---|---|
| Backup restore (Cloud SQL) | Quarterly |
| Redis failover test | Semi-annually |
| Full DR tabletop exercise | Annually |
| Full DR live failover | Annually |
Next: Risk Assessment โ