Acoustic Transmission Loss in Healthcare Facilities

By Sarah Okonkwo · February 25, 2026

Acoustic Transmission Loss in Healthcare Facilities

1) Introduction: What you’ll learn and why it matters

Healthcare buildings are full of sound sources that are hard to “turn down”: nurse stations, rolling carts, alarms, MRI chillers, paging systems, visitors, and HVAC. At the same time, the stakes are high—sleep, recovery, speech privacy, and staff performance are all affected by noise. Acoustic transmission loss (TL) is one of the most practical tools you can use to control how much sound passes through walls, doors, ceilings, glazing, and service penetrations.

This tutorial shows you how to evaluate existing partitions, measure field performance, identify weak links (doors, flanking paths, penetrations), and specify fixes. You’ll leave with a repeatable workflow you can use on real projects: patient rooms, consultation rooms, exam rooms, behavioral health units, and corridors adjacent to nurse stations.

2) Prerequisites / setup requirements

Test gear: Class 1 or good Class 2 sound level meter (SLM) with 1/3-octave bands; acoustic calibrator (94 dB at 1 kHz); an omnidirectional source (dodecahedron speaker preferred) or a powered speaker capable of steady pink noise; measurement mic(s) if using software.
Software (optional but helpful): Smaart, Room EQ Wizard (REW), or similar for RTA/1-3 octave logging.
Signal: Pink noise or 1/3-octave band-limited noise. Target source levels should exceed background noise by at least 10 dB in each band you care about (more is better).
Access: Ability to use two adjacent rooms (source room and receiving room) and to temporarily control door positions, HVAC modes (if possible), and occupancy.
Context knowledge: Familiarity with STC (lab rating) versus field performance (often reported as FSTC or ASTC). In healthcare, you’ll also be thinking about speech privacy and audibility of alarms.

3) Step-by-step workflow

Step 1 — Identify the real-world acoustic problem and set a target

Action: Write a one-sentence goal tied to a space and a source. Example: “Reduce speech transmission from Corridor A into Patient Room 214 so normal conversation is unintelligible at the bed.”

Why: TL is frequency-dependent and source-dependent. A wall that performs fine for HVAC rumble may fail for speech intelligibility (500 Hz–4 kHz). A clear goal keeps you from overbuilding the wrong element.

Specific targets you can use:
- Consultation / interview rooms: Aim for partitions/door assemblies that deliver field performance comparable to STC 45–55 depending on privacy needs.
- Typical patient room to corridor: Many projects target around STC 45+ for wall/door assemblies, but actual field results often come in lower if doors and penetrations aren’t treated.
- Behavioral health: Higher isolation is common, but ensure solutions align with safety and ligature requirements.
Pitfalls: Choosing a target based only on wall STC and ignoring the door, glazing, or above-ceiling paths. In a hospital, the door is frequently the dominant leak even when walls are well-built.
Step 2 — Survey the construction and mark likely weak points

Action: Walk the boundary between the two spaces and document: wall type, door type, glazing, ceiling plenum continuity, duct penetrations, medical gas penetrations, back-to-back electrical boxes, and any gaps.

Why: Transmission loss in the field is commonly limited by flanking and leakage, not the nominal wall layers. A single 1% open area can destroy isolation at mid/high frequencies.

What to look for (common in healthcare):
- Door undercuts: A 10–20 mm undercut is a major speech leak. Measure it.
- Frame gasketing: Missing, damaged, or paint-sealed door seals.
- Above-ceiling bypass: Wall stops at ceiling grid while plenum continues over it (classic corridor-to-room flanking).
- Return air paths: Transfer grilles or door grilles short-circuiting isolation.
- Headwall penetrations: Medical gas/electrical penetrations around patient headwalls that weren’t sealed.
Pitfalls: Assuming “thicker wall = better” while ignoring that the wall may not be full height to the deck. Another common miss is overlooking that two rooms share ductwork with no lined offsets, carrying speech-like noise.
Step 3 — Calibrate your measurement chain and log background noise

Action: Calibrate the SLM/mic with a 94 dB @ 1 kHz calibrator before and after the session. Then measure background noise in both rooms (HVAC on, typical operating condition) in 1/3-octave bands for at least 30 seconds of averaging.

Why: Field TL measurements are only meaningful if you know whether background noise is masking the received signal. Hospitals often have high steady HVAC noise plus intermittent alarms, which can contaminate results.

Numbers to use:
- Target ≥ 10 dB signal-over-noise in the receiving room per band; ≥ 15 dB is safer.
- If background is high at 125–250 Hz due to air handlers, plan to focus on 500 Hz–4 kHz for speech-driven goals, or schedule testing at quieter times.
Pitfalls: Forgetting post-calibration (you won’t know if the mic drifted). Measuring background with doors propped open “for convenience,” which invalidates later comparisons.

Troubleshooting: If you can’t get 10 dB over background, raise the source level (within safety limits), move the source away from boundaries to reduce modal nulls, or choose band-limited noise centered where you need data.
Step 4 — Set up the source correctly (this is where many tests go wrong)

Action: Place the speaker/dodecahedron in the source room at least 1.0 m from the test partition and 1.0 m from major side walls if possible. Play pink noise and adjust to achieve an average level of about 85–95 dB SPL in the source room (measured at multiple mic positions).

Why: You’re trying to excite the room broadly and avoid a measurement dominated by one standing wave or one boundary reflection. In hospitals, rooms can be small and reflective; placement matters.

Technique: Use at least 3 source-room mic positions (ear height ~1.2–1.5 m) and average. If using an SLM, log Leq 10–20 seconds per position.

Pitfalls: Placing the source right against the wall (artificially increases coupling at some bands). Setting levels too low (received signal disappears into HVAC). Setting levels too high (distortion or staff complaints).

Troubleshooting: If you see erratic low-frequency results, move the source and/or mic positions and increase averaging. If the speaker is distorting, reduce level or use a larger source.
Step 5 — Measure received levels and compute a practical “in-field TL” snapshot

Action: In the receiving room, measure 1/3-octave SPL at 4–6 mic positions, again at ear height, avoiding corners (stay >0.5 m from walls). Record the average received spectrum. Compute the simple level difference per band: D(f) = L_source(f) − L_receive(f).

Why: The raw level difference D(f) quickly shows where the assembly is weak. Even before doing full ASTM-style normalization, you’ll see if the problem is a high-frequency leak (gasketing) or mid-band flanking (plenum/duct).

Settings: Use 1/3-octave with slow averaging. Capture at least 10 seconds Leq per position. If alarms occur, discard those segments and re-measure.

Common patterns and what they mean:
- D collapses above 1 kHz: door gap, glazing leak, unsealed penetration.
- D poor at 125–250 Hz but decent above 500 Hz: structural/flanking through slab, lightweight wall resonance, or HVAC/duct coupling.
- D “notch” around 500–800 Hz: cavity resonance, back-to-back boxes, or a specific flanking element.
Pitfalls: Measuring with different door positions between source and receive measurements. Using only one mic position (you may land in a room null and misdiagnose the assembly).

Troubleshooting: If received levels don’t change much when you increase the source by 10 dB, you’re likely limited by background noise or you’re not actually exciting the partition (check source routing/level, confirm the noise is playing).
Step 6 — Find and confirm flanking paths with a “seal-and-test” approach

Action: Temporarily treat suspected leaks one at a time and repeat a quick measurement (even a single mic position can show trends). Useful temporary tools: painter’s tape for small gaps, door sweep temporarily taped in place, blankets over transfer grilles (only for short tests), and removable putty around penetrations.

Why: Hospitals rarely fail because a wall’s gypsum layers are wrong; they fail because sound bypasses the wall. You want proof of which path dominates before recommending construction changes.

What to try (in order):
- Door perimeter: Tape paper strips around the door and look for airflow/sound leaks; temporarily add compression with taped foam.
- Undercut: Temporarily block with a draft stopper; watch changes above 1 kHz.
- Ceiling plenum: If there’s a lay-in ceiling, temporarily block around return openings or ceiling gaps; check mid/high changes.
- Penetrations: Seal around conduit/pipe entries on both sides and retest.
Pitfalls: Confusing “improvement at one frequency” with a full fix. Also, never leave airflow paths blocked in an operational healthcare space—keep these tests brief and coordinated with facilities staff.

Troubleshooting: If sealing the door doesn’t change results, you may be flanking above the ceiling or through ductwork. If sealing penetrations doesn’t help, look for shared returns or continuous plenum.
Step 7 — Specify upgrades that match the failure mode (and can actually be built)

Action: Choose interventions based on what your measurements revealed, prioritizing the biggest leaks first.

Why: TL improvements are not additive in the way people expect. A wall that could be STC 55 won’t behave like it if a door performs like STC 25. Your best ROI usually comes from fixing the weakest element.

Common upgrades with concrete details:
- Door system (most common fix): Replace hollow-core with a solid-core 45–60 mm door; add continuous perimeter seals and an automatic door bottom (drop seal). Target <3 mm effective gaps around the perimeter.
- Wall to deck: Extend partitions to structure (slab/deck) and seal the top track with acoustical sealant. If that’s impossible, add a tested plenum barrier system; don’t assume batts alone solve flanking.
- Penetration sealing: Use approved acoustical/firestop systems; seal both sides. Pay attention to back-to-back boxes—offset them by at least 600 mm horizontally or use putty pads and sealed boxes.
- Glazing: If a vision panel is required, specify laminated glass and ensure frame seals are continuous. A high-STC wall with a poorly sealed vision lite will underperform.
- Duct/return paths: Add lined duct offsets, plenums with acoustic lining, or transfer silencers. A straight, shared return can carry speech remarkably well.
Pitfalls: Specifying a “high-STC wall” without a matching door rating and installation details. Another frequent issue is forgetting infection control: materials and details must be cleanable and compliant (avoid solutions that trap dust or can’t be wiped down).
Step 8 — Re-test and document improvement with the same method

Action: Repeat Steps 3–5 with the same source position, mic heights, door positions, and HVAC mode. Create a before/after plot of D(f) from 125 Hz to 4 kHz, and summarize changes at key bands (500 Hz, 1 kHz, 2 kHz).

Why: Healthcare stakeholders respond to evidence. A clean before/after comparison also protects you: it shows which intervention produced which improvement.

Pitfalls: Changing multiple variables at once (new source level, different mic positions, different background conditions). If you must change conditions, note it explicitly and avoid overstating conclusions.

4) Before and after: what results should look like

Example corridor-to-patient-room scenario (speech privacy issue):

Before: D(f) around 18–22 dB at 1–2 kHz, with a noticeable drop above 2 kHz. Staff conversations in the corridor are clearly intelligible at the bed, especially at night when patients are trying to sleep.
After door sealing + automatic drop seal: D(f) improves to 28–35 dB at 1–2 kHz, and high-frequency isolation no longer collapses. Corridor speech becomes mostly unintelligible unless someone speaks loudly right outside the door.

Expected qualitative change: less “sibilant spill” (the sharp S/T consonants) through the door gaps, fewer complaints about privacy, and improved patient rest—without having to rebuild the wall.

5) Pro tips for taking it further

Use a speech-weighted check: After your pink-noise work, play standardized speech material at a realistic level (~60–65 dBA at 1 m) in the corridor and measure received dBA at the bed. It connects your TL work to what people actually perceive.
Don’t ignore ceilings: In many hospitals, the wall is fine but the ceiling plenum is a highway. If walls don’t go to deck, treat the plenum path explicitly with a tested barrier and sealed edges.
Track the “weakest link” mathematically: If a door leaks, the effective isolation is dominated by that path. Even small gap reductions can outperform adding another gypsum layer to the wall.
Coordinate with clinical operations: Plan tests around shift changes, cleaning rounds, and quiet hours. Capture notes about intermittent alarms; otherwise your data will look inconsistent.
Validate workmanship: The best spec fails with poor installation. Bring a flashlight: you can often see light around door seals or through penetrations—if light passes, sound passes.

6) Wrap-up: build the habit of measuring, fixing, and confirming

Transmission loss work in healthcare is rarely glamorous, but it’s one of the most impactful skills you can develop as an audio practitioner working in the built environment. Measure the problem in bands, find the dominant leak, fix the simplest path first, and re-test under the same conditions. Do that consistently and your recommendations will start landing as predictable improvements instead of hopeful upgrades.

Run this workflow on a single door-and-wall boundary in your next facility walk-through. Even one well-documented before/after will sharpen your instincts for the next ten rooms.