World Sleep Industry ENCYCLOPEDIA

Portable Polysomnography Recorders

2.2.3

Definition, Portability Boundary, and Study Type

A portable polysomnography recorder is a medical recording system that travels with the patient and continuously acquires and retains multiple sleep-related physiologic signals using a battery or portable power source. It commonly concentrates patient-side amplification, sensor interfaces, analog-to-digital conversion, a clock, and local storage in a worn recorder; professional software retrieves, scores, and interprets the study afterward. “Portable” describes size, power, transport, and use environment. It does not mean that any person may self-apply the device or that the recording automatically produces a diagnosis.

In a common classification, an unattended full portable PSG that records EEG, EOG, chin EMG, ECG or heart rate, airflow, thoracic and abdominal effort, and SpO2 and thereby supports sleep staging approximates a Type II or Level 2 study. A Type III study with limited respiratory, cardiac, and oxygenation channels is chiefly directed at a defined sleep-breathing question; systems with fewer channels, stand-alone oximetry, and consumer sleep trackers occupy other levels. The words “multichannel,” “home,” or “professional” do not replace inspection of the actual channels and intended use.

One recorder can have basic and extended kits, adult and pediatric accessories, professional-applied and self-applied modes, or separate Level 2 and Level 3 configurations. The same host device therefore does not confer the same capability on every recording. The report and equipment record should retain the recorder, firmware, software, lead set, sensors, kit mode, and configuration version used that night. A family's highest specification must not describe a study that in fact lacked sleep-staging channels.

Patient-Side Architecture, Channel Configuration, and Continuous Recording

A typical system includes a worn recorder, belt or clip, patient cables, EEG, EOG, and EMG electrode inputs, ECG, a nasal-pressure cannula and oronasal thermal airflow sensor, thoracic and abdominal effort belts, pulse oximetry, position and snoring sensors, and optional leg EMG, carbon dioxide, or auxiliary inputs. The recorder may retain every waveform locally while also sending status or data to a phone, access point, or cloud service. Wireless communication is an additional path and should not be the sole protection for record integrity unless the system was explicitly designed and validated that way.

Every channel needs continuous writing against a common, stable time base. Sensor insertion, Bluetooth reconnection, restart, daylight-saving changes, or a phone clock adjustment must not create an undisclosed jump, repetition, or misalignment. The system should distinguish acquired measurements, held last values, automatic interpolation, algorithm-derived values, and manual event marks and retain logs for start, pause, reconnection, battery, storage, errors, and configuration. A device that retains only summary indices without reviewable waveforms does not meet the review needs of a complete multichannel record.

Recorder mass, attachment site, and cable routing influence sleep posture, sensor traction, and actual signal quality. Placement should avoid a rolled belt, lateral pressure point, and sweat or condensation at connectors. Leads need enough slack for movement without creating neck entanglement, trip, or snag hazards. Battery capacity and storage must cover preparation, delayed sleep onset, a whole-night study, and a reasonable margin. A claimed maximum duration is meaningful only with the stated channel load, wireless mode, battery condition, environment, and storage configuration.

Application Responsibility, Self-Use, and Start Confirmation

Hookup may be performed by a technologist at a facility before the patient goes home, by a technologist in the home, or by the patient and caregiver from instructions. These models contain different critical tasks and require separate validation. Successful professional application does not show that a lay user will place electrodes, tension effort belts, and recognize channel status correctly. One successful training session also does not represent people with different vision, dexterity, language, skin, body shape, or cognitive ability. Labeling should specify permitted operators, ages, settings, and assistance.

A self-application kit needs packaging in task order, connections keyed by unambiguous color, shape, or geometry, mirror-oriented placement illustrations, and actionable feedback for laterality, vertical location, tension, skin preparation, and electrode contact. An app or display should name a channel and a correctable cause such as disconnected, poor contact, no pulse, blocked cannula, or insufficient storage rather than show only an unexplained red lamp. A video, helpline, or remote assistant can supplement instructions, but the system still needs a complete offline path when the network is unavailable.

Before initiation, the workflow verifies the patient and order, recorder identity, time, battery, storage, kit, disposable supplies, and scheduled start and stop conditions and performs impedance or signal checks and biocalibrations appropriate to the configuration. While awake, the patient practices lying down, turning, getting out of bed, refixing only the components that users are permitted to handle, and ending the record. The person should not be expected to improvise electrode relocation, open the recorder, or reset clinical parameters. An automatic start also needs a clear confirmation so the device does not remain in standby all night or exhaust its battery during an unintended early recording.

Unattended Quality Control, Failure, and Recovery

The central unattended limitation is that a technologist cannot immediately repair a critical detached sensor. A lifted frontal electrode, failed reference, displaced pressure cannula, misaligned thermal sensor, loose effort belt, compressed oximeter, shifted recorder, or cable pulled by a pet may persist for hours without waking the patient. The device should perform an initial completeness check and log channel quality without unnecessarily disturbing sleep. If remote status viewing is used, the service must define who watches, when intervention occurs, and what responsibility remains after connectivity is lost.

Redundancy can preserve analyzability—for example, simultaneous nasal-pressure and oronasal thermal airflow, thoracic and abdominal effort, or an alternate EEG derivation—but a substitute is not automatically equivalent. The system should report valid minutes for every channel, artifact type, packet loss and restart intervals, common-time offset, and any substitution rule. If EEG, EOG, or chin EMG is inadequate for reliable staging, the recorder's nominal channel count must not keep the study classified as complete PSG. Missing airflow, effort, or oxygenation likewise must not be silently model-filled and used for definitive respiratory classification.

Technical acceptability is judged against the ordered question, not merely whether a file opens. Before deployment, professionals define the minimum valid channels, coverage duration, and required stages or positions; after return, they review raw waveforms channel by channel. When data are insufficient, the failure mode can lead to reconnection, retraining, accessory replacement, another home recording, or an attended study. Multinight recording can expose night-to-night variation and increase the chance of a successful night, but it cannot cancel a repeated systematic misapplication, electrode error, or algorithmic bias.

Clinical Use, Scoring Interpretation, and the Non-Alarm Boundary

Portable full PSG can record habitual bedtime, bed partner, and usual position in a familiar setting and may reduce travel burden. It also lacks continuous technologist observation, immediate sensor repair, on-site adjustment of the video view, and protocol-driven intervention. Sleep architecture or event indices may differ between home and laboratory because of environment, schedule, night-to-night variation, hookup, and signal quality. Not every difference proves that home is more natural or one device more accurate, and home recording does not guarantee elimination of a first-night effect.

Study selection follows clinical assessment, suspected disorder, patient capability, and local rules. Selected adults with suspected uncomplicated OSA may be served by a more limited HSAT. When reliable stage, arousal, limb movement, behavior, or multisystem relationships are needed, full portable PSG can supply more information but does not always replace attendance. A patient who may need immediate response, complex PAP titration, expanded seizure electrodes, precise video-behavior observation, or cannot safely manage the equipment is more likely to need professional on-site capability.

Scoring uses the declared rule set and actual valid signals, with review by qualified personnel and interpretation by an appropriate clinician. The recorder, app, or cloud platform may prestage sleep and premark events, but a risk score, AHI, or hypnogram does not directly diagnose every sleep disorder. Most portable recorders are also not vital-sign alarm monitors or automatic apnea monitors. Unless a specific model has that intended use and supporting validation, an abnormal overnight SpO2, heart rate, or respiration is not guaranteed to be detected in real time, and the recorder cannot carry emergency-watch responsibility.

Home Safety, Data Handoff, and Turnaround Maintenance

The home may contain unstable outlets, extension cords, liquids, children, pets, carpet static, low lighting, temperature and humidity variation, and no continuous professional support. Isolation, leakage current, electromagnetic compatibility, drops and vibration, fluid ingress, batteries, charging, and essential performance should be verified in the intended environment. Labeling must state whether mains power, a power bank, or a phone connection is permitted during recording. Abnormal heat, swelling, odor, liquid entry, cracking, repeated restart, or lead damage requires discontinuation; the recorder must not continue operating or charging beneath a pillow or bedding.

Skin preparation, adhesives, conductive materials, and fixation pressure should suit overnight contact. Reusable belts, leads, probes, and recorders are cleaned, disinfected, and dried by validated methods; a cannula, electrode, or other single-use item is not improvised for reuse. On return, the service checks components, sharp edges, tensile damage, loose connectors, battery health, memory, clock, and cleanliness and quarantines failed accessories. The transport case also keeps electrode paste, liquid, and contaminants out of electronic connectors, and equipment exposed to temperature extremes must meet recovery conditions before use.

Patient identity, raw waveforms, logs, and reports must remain linked through configuration, issue, return, upload, scoring, and archive. Loss or shipping error must not expose the prior user's data to the next user. Local storage and transfer need access control, encryption, integrity verification, auditability, and secure deletion. When wireless or portal service fails, the design should define offline recording, buffer capacity, deduplicated retransmission, and manual recovery. A progress bar alone does not establish successful upload; file checks, channel duration, and readability do.

Acceptance Metrics, Service Capability, and Direction

Acceptance covers bench performance, an overnight simulation, representative intended users, and the actual use environment. Beyond gain, noise, sampling, filtering, clock behavior, saturation recovery, and physiologic-channel performance, it tests the complete journey from issue and application through start, nighttime mobility, stop, return, and data import. Measures include first-attempt success, valid fraction by channel, simultaneous valid time for critical signals, calls for help and use errors, repeat rate, battery and storage margin, synchronization error, transfer failure, cleaning turnaround, and report traceability, stratified by professional and self-application and by relevant populations.

The service maintains asset and configuration records, accessory life, charging cycles and battery replacement, calibration and functional checks, cleaning, software and security updates, fault root causes, spare units, and recall procedures. A failed study must not automatically be blamed on the patient; instructions, training, supplies, connectors, quality-detection logic, transport, and hardware aging are distinct causes. After software or accessory changes, representative records should confirm legacy readability, continuity of results, and preservation of critical user tasks.

Development directions include dry electrodes with fewer leads, wireless or patch-based channels, guided setup, edge quality detection, remote support, multinight recording, and explainable assisted scoring. Less mass and cabling may improve acceptance but also changes montage coverage, reference method, battery, clock synchronization, and detachment modes. Artificial intelligence can flag a suspect interval but cannot reconstruct physiology that was never acquired. Progress means demonstrating a higher usable-data yield, fewer repeat studies, and clinically concordant results in a defined population—not merely shrinking laboratory equipment. Related entries include polysomnography systems, portable sleep respiratory monitors, home sleep screening devices, and sleep-test data analysis software.