Difficulty: Intermediate | Prerequisites: General understanding of the U.S. healthcare system; familiarity with basic psychology concepts (bias, cognition).
This chapter sits at the intersection of healthcare delivery and human psychology. It covers how the U.S. healthcare system is structured (the insurance plans patients navigate), how providers and patients communicate (and where that communication breaks down), and the emerging role of artificial intelligence in clinical settings. If you understand the basics of cognitive biases from introductory psychology, you already have a foundation here. The chapter matters because poor communication between providers and patients directly affects treatment outcomes, adherence and patient safety.
The U.S. healthcare system uses several insurance models (HMO, PPO, POS, indemnity) that shape how patients access care. Communication failures on both the provider and patient side, from jargon to cognitive biases, reduce treatment adherence and erode trust. AI shows diagnostic promise but raises concerns about transparency, patient trust and data privacy.
HMO (Health Maintenance Organisation)
A managed-care plan where members choose a primary-care physician from a fixed pool and pay a small co-pay per visit. Specialist visits and non-emergency hospital stays typically require pre-approval. Think of it as a "stay in the network" plan.
PPO (Preferred-Provider Organisation)
A network of doctors offering discounted rates to plan members. Patients usually do not need prior authorisation to see an in-network specialist. In simple terms, you get more flexibility than an HMO, but your costs are lowest when you stay in-network.
POS (Point-of-Service Plan)
A hybrid administered by an insurance company or HMO that lets members go out of network for an additional cost. Members usually need a referral to see a specialist. Think of it as an HMO with an escape hatch.
Traditional Indemnity Plan
Patients choose their own doctors and hospitals and pay on a fee-for-service basis. No referrals needed. In simple terms, maximum freedom, maximum out-of-pocket cost.
Jargon (in clinical communication)
Specialised medical language used by providers that can prevent patients from asking questions. About 90% of patients report they do not prefer jargon.
Baby Talk
Overly simplistic explanations from a provider. The opposite extreme of jargon, and equally unhelpful.
Elder Speak
Short, simplistic speech directed at older patients. Can produce refusal to care and undermine self-esteem.
Nonperson Treatment
Depersonalisation of patients by providers as a form of emotional self-protection. In simple terms, treating a patient as a case rather than a person.
MBI (Maslach Burnout Inventory)
A tool that measures provider burnout. Research using the MBI shows that providers use nonperson treatment to protect themselves emotionally.
Objectivity Illusion
The tendency to overestimate the extent to which one's personal views are widely shared. Think of it as assuming your perspective is the default.
Bias Blind Spot
Being unaware of your own biases while believing you are less biased than others.
False Consensus Effect
Believing everyone shares the same judgement or view as you.
Self-Serving Bias
Attributing your own mistakes to circumstances while attributing other people's mistakes to their personal faults or character.
Confirmation Bias
Searching for, interpreting, recalling or accepting information that confirms an existing belief, while ignoring information that challenges it.
Model Overfitting (in AI)
When a model performs well on training data but poorly on new, unseen data because it has become too specific to the training examples. In simple terms, the model memorised the answers instead of learning the principles.
Black Box Issue (in AI)
The inability to see how or why an AI algorithm produces a particular treatment recommendation. The reasoning is hidden inside the model.
Placebo Effect
Any medical procedure that produces an effect in a patient because of its therapeutic intent rather than its specific nature. A medical setting with trusted providers tends to produce stronger placebo effects.
Four main insurance structures determine how patients access providers:
HMO – Choose a primary-care physician from the plan's pool. Small fixed co-pay per visit. Specialist referrals and non-emergency hospital visits need pre-approval.
PPO – A network of providers at discounted rates. No prior authorisation needed for in-network specialists. More flexibility, higher premiums.
POS – A hybrid that allows out-of-network visits at extra cost. Referrals are usually required to see a specialist.
Traditional Indemnity – Fee-for-service. Patients pick any doctor or hospital. No referrals required. Highest out-of-pocket costs.
The key distinction across these plans is the trade-off between freedom of choice and cost control.
Inattentiveness – Not listening to patients or cutting them off before they finish describing concerns. The provider jumps to a diagnosis too early.
Norms – Talking about communication scores encourages providers to improve, suggesting accountability structures can shift behaviour.
Computer use – High computer use by clinicians in safety-net clinics is associated with lower patient satisfaction and observable communication differences (Lakhlifi et al., 2023).
Jargon – Specialised language that discourages patients from asking questions. Roughly 90% of patients do not prefer jargon.
Baby talk – Overly simplistic explanations that can feel patronising.
Elder speak – Short, simplistic speech targeted at older people. Can cause patients to refuse care (Schroyen et al., 2018).
Nonperson treatment – Depersonalising patients as an emotional defence mechanism. The MBI (Maslach Burnout Inventory) shows providers use this to cope with burnout.
Stereotyping – Cuts across all patient identities and affects both treatment quality and care. Research shows stereotypes lead providers to assume Black patients experience less pain than White patients.
Lakhlifi et al. (2023) – Found that clinicians had poor statistical proficiency combined with high confidence, even when their answers were incorrect.
Patient-side barriers to effective communication include:
Lack of education or health literacy
Inability to present symptoms clearly
Inattention driven by unmanaged anxiety
Key research:
Schroyen et al. (2018) – Elderspeak leads to refusal of care and increases in self-esteem and confidence (patients push back). Physicians and medical students adapt their speech when explaining treatments to older patients.
Redelmeier & Ross (2018; 2019) – Medical care regularly includes situations where a colleague says something the responsible clinician does not want to hear.
Cognitive biases that affect both patients and providers:
Objectivity illusion – Overestimating how widely your personal views are shared.
Bias blind spot – Being unaware of your own bias while believing you are less biased than others.
False consensus effect – Assuming everyone holds the same judgement as you.
Self-serving bias – Attributing your own lapses to circumstances and other people's lapses to their character.
Confirmation bias – Seeking, interpreting and recalling information that confirms what you already believe.
Recommended corrective: Show greater humility, consider colleagues' views thoughtfully, and remain open to changing perspectives.
When provider-patient communication breaks down, several outcomes follow:
What decreases:
Trust
Treatment adherence
Follow-up care
Consistent care
What increases:
Filing of formal complaints
Patients seeking emotionally satisfying treatment (comfort over clinical effectiveness)
Potential benefits:
Accuracy and speed in analysing medical images
Enhanced diagnostics
Concerns:
Privacy – Risk of data breaches and personal information being leaked.
Training – AI learns from large datasets of medical information, which may contain biases.
Black box issue – Clinicians and patients cannot see how or why the algorithm reached a recommendation.
Not person-centred – Current AI is not designed around patient-centred care.
Low patient trust – Patients generally do not trust AI for medical decisions.
Model overfitting – The model performs well on training data but poorly on unseen data because it has become too specific to the examples it learned from.
The 3 Cs of healthcare roles:
Cure – The physician's responsibility. Performing any treatment action with the potential to restore patients to good health.
Care – The nursing staff's orientation. The humanitarian side of medicine: keeping patients' physical and emotional state in balance.
Core – Hospital administration. Ensures smooth functioning of the system and flow of resources.
Communication among hospital staff:
High occupational segregation exists between different roles in the hospital. Different professional groups (physicians, nurses, administrators) often operate in silos.
Burnout:
Marked by emotional exhaustion, cynicism and a low sense of self-efficacy on the job.
Linked to absenteeism, high job turnover, lengthy work breaks and, in severe cases, suicide.
Irving Janis's landmark study:
Changed how patients are prepared for surgery. The study examined how pre-surgical fear levels affected post-operative outcomes: highly fearful patients and patients with very little fear both had worse outcomes after surgery, suggesting moderate, well-managed anxiety leads to better preparation and recovery.
CAM therapies:
Yoga – Combines breathing techniques, posture, strengthening exercises and meditation. People report lower stress and anxiety.
Chiropractic medicine – Adjustments on the spine and joints to correct misalignments. Thought to prevent and cure illness. About 20% of Americans use chiropractic services.
Overall evaluation – There is insufficient evidence to evaluate the effectiveness of CAM therapies as a whole.
Placebo effects:
A placebo effect occurs when a medical procedure produces an effect because of its therapeutic intent, not its specific nature. The setting matters: a medical environment with trusted, confident providers will produce stronger placebo effects.
Students often confuse the bias blind spot with the false consensus effect. The blind spot is about believing you are less biased than others. The false consensus effect is about believing others share your views.
Students sometimes think jargon and baby talk are the same problem. They are opposite ends of the same spectrum: one is too complex, the other too simple.
Students may assume AI in healthcare is already person-centred. It is not. Current AI models are not designed around patient-centred care principles.
Students often overlook that elderspeak can cause patients to refuse care, not just feel patronised.
Know the four healthcare plan types and what distinguishes each (referrals, pre-authorisation, cost structure).
Be able to list and define each cognitive bias (objectivity illusion, bias blind spot, false consensus effect, self-serving bias, confirmation bias) with a clinical example.
Understand the consequences of poor communication: what goes up and what goes down.
Know the 3 Cs (Cure, Care, Core) and which professional role each maps to.
Be prepared for questions about AI limitations: black box issue, overfitting, patient trust.
Know the key researchers and their findings: Schroyen et al. (2018), Redelmeier & Ross (2018; 2019), Lakhlifi et al. (2023).
True or False: In a PPO, patients always need prior authorisation to see a specialist. (False – PPOs typically do not require prior authorisation for in-network specialists.)
Fill in the blank: The tendency to attribute your own mistakes to circumstances and others' mistakes to their character is called ________. (Self-serving bias.)
True or False: About 90% of patients prefer when their doctor uses medical jargon. (False – 90% do not prefer jargon.)
Fill in the blank: The three Cs of healthcare are Cure, Care and ________. (Core.)
True or False: AI in healthcare is currently considered person-centred. (False.)
Q: What is the key difference between an HMO and a PPO?
A: An HMO requires patients to choose from a fixed pool of physicians and get pre-approval for specialists, while a PPO offers a broader network with no prior authorisation needed for in-network specialists.
Q: Name three cognitive biases discussed in this chapter and explain how each could affect clinical care.
A: Confirmation bias (a provider looks for evidence that supports their initial diagnosis and ignores contradictory signs), self-serving bias (a provider attributes a poor outcome to the patient's non-compliance rather than their own diagnostic error), and bias blind spot (a provider believes they are less biased than their colleagues).
Q: What did Schroyen et al. (2018) find about elderspeak?
A: Elderspeak results in patients refusing care and can increase self-esteem and confidence (they push back). Physicians and medical students were found to adapt their speech when explaining treatments to older patients.
Q: What is the "black box issue" in AI healthcare, and why does it matter?
A: The black box issue refers to the fact that clinicians and patients cannot see how or why an AI algorithm produces a particular treatment recommendation. This matters because it undermines trust and makes it difficult to verify or challenge the recommendation.
Q: What are the consequences of poor provider-patient communication?
A: Trust, treatment adherence, follow-up care and consistent care all decrease. Complaints and patients seeking emotionally satisfying (rather than clinically optimal) treatment both increase.
This material connects directly to Chapter 10 (Pain Management), where provider communication and patient trust affect how pain is reported and treated. The cognitive biases covered here reappear in Chapter 11 (Chronic Health Disorders) when discussing how patients adjust to long-term illness and how providers' assumptions shape quality of life. The burnout and nonperson treatment concepts also tie into the broader field of organisational psychology and occupational health.
HMO, PPO, POS, indemnity plan, fee-for-service, managed care, provider communication, jargon, baby talk, elderspeak, nonperson treatment, depersonalisation, MBI, Maslach Burnout Inventory, burnout, stereotyping, racial bias in pain, cognitive bias, objectivity illusion, bias blind spot, false consensus effect, self-serving bias, confirmation bias, AI in medicine, black box problem, model overfitting, patient trust, CAM, complementary and alternative medicine, yoga, chiropractic, placebo effect, 3 Cs, cure care core, Irving Janis, surgical preparation, Schroyen 2018, Redelmeier Ross, Lakhlifi 2023, PSY 320, health psychology, Michigan State University