WHO defined primary care “as socially appropriate, universally accessible, scientifically sound first level care provided by a suitably trained workforce supported by integrated referral systems and in a way that gives priority to those most needed, maximizes community and individual self-reliance and participation and involves collaboration with other sectors, including health promotion, illness prevention, care of the sick, advocacy and community development” [1].

Therefore, primary care services are at the integrated people-centered healthcare in many countries where they provide an entry point into the health system and have directly impact on people’s well-being and their use of other health and care resources. Unsafe or ineffective primary care may increase morbidity and preventable mortality, and may lead to the unnecessary hospitalization and specialist resource and, in some cases, disability and even death [2].

Patient safety is the absence of preventable harm to a patient during the process of healthcare and reduction of risk of unnecessary harm associated with healthcare to an acceptable minimum. An acceptable minimum refers to the collective notions of given current knowledge, resources available and the context in which care was delivered weighed against the risk of non-treatment or other treatment [2]. Therefore, it is the minimum prerequisite for high-quality care.

Studies on patient safety have traditionally focused on hospital care. The reason for this is that it reflects the dominance of hospital-based care in many health systems and is the result of a perception that this is where most serious incidents occur.

Primary care has been perceived for many years as a low technology environment where safety would not be a problem. However, in England, 90% of contacts with the National Health Service take place in primary care, and more than 750,000 patients consult their GP each day. In many countries, it is estimated that 85% of all healthcare contacts occur in primary care [1].

European data show that the issue of patient safety is ongoing and that, for example, in the United Kingdom between 5 and 80 safety incidents occur per 100,000 primary care consultations, which translates to between 370 and 600 incidents per day [3].

For this reason, patient safety and preventable harm to patients in primary care is becoming a rising issue because it represents the largest volume of healthcare encounters.

Attention towards patient safety was renewed in 2016 by the World Health Organisation (WHO) with its “Technical Series on Safer Primary Care” aiming at raising awareness about the underlying causes of safety incidents and consequences of unsafe primary care [2].

There is lack of a formal reporting mechanism for medical errors in primary care. Incident reporting is practiced as a self-reporting process and the magnitude of errors could have been underestimated. Among existing strategies to help improve clinical effectiveness and enhance patient safety there are the Quality and Outcomes Framework (QOF), appraisal, revalidation, significant event analysis (SEA), and critical incident reporting systems (CIRS) One of the first comprehensive and coordinated attempts to improve patient safety in primary care is the Scottish Patient Safety Program in Primary Care (SPSP-PC), established in March 2013.

Recent years have seen more researches located in primary care settings which have different features compared to secondary care. Attempts to classify medical errors and preventable adverse events in primary care have proved challenging due to the lack of an evidence base and are yet to be reliably quantified. Data on the most frequent misdiagnosed conditions are scarce, and little is known about which diagnostic processes are most vulnerable to breakdown. Most of them are derived from studies of malpractice claims or self-report surveys. These methods introduce significant biases that limit the generalizability of findings to routine clinical practice [4].

Many countries have implemented strategies to reduce avoidable harm or “never event” defined as “a serious, largely preventable patient safety incident that should not occur if the available preventable measures were implemented by healthcare workers.” In 2014, De Wet published a never event list based on general practice identified eight items (Mistaken patient identity, Acts of omission, Investigations, Medication, Medico-legal and ethical incidents, Clinical management Practice systems, Teamwork and communication) and, if there is some evidence of reduction of patient safety incidents in some item of never event list, it is unclear whether all of the never events in the list are truly preventable, or which of the available interventions will be acceptable or effective [5, 6].

The difficulties in identifying and monitoring the never event list is due to the specific context of General Practice.

“Marshall Marinker has characterised the role of the GP as being to ‘marginalise danger’, contrasting this with the specialist, whose diagnostic role is to ‘marginalise uncertainty’. In other words GPs have the often difficult task of identifying the minority of patients whose presenting symptoms represent serious illness from the majority who do not have something seriously wrong” [7].

Patients in primary care setting can have many health problems (multimorbidity), complex needs (both social and medical), and frequent interactions with healthcare staff in a number of different clinical contexts. There is a range of challenges for GPs, due to the specific primary care setting based on deliver optimal disease management and patient-centered care in a time-limited consultation. The many and varied diseases encountered in primary care make comprehensive measurement of guideline adherence difficult (especially for guidelines that change frequently). Decision-making in primary care often relies on complicated care algorithms specific to numerous diseases. The complexity and inadequacy of single disease guidelines, evidence-based medicine and barriers to shared decision-making were managed through the use of relational continuity of care.

This peculiar care relationship involves a many patient factors, including sex, age, the nature of the illness, earlier experiences and perceived control on the illness, education, financial considerations, personal values, and cultures and traditions [8].

Studies suggests inappropriate care with patients presenting the same conditions, as a result of gender, ethnicity, or socioeconomic disparities and some vulnerable social groups are more likely to experience adverse patient safety events. A recent research confirms that, in primary care, women and black patients are more likely to receive inappropriate diagnosis, treatment, or referrals compared to men and Whites, respectively. However, our findings interestingly suggest that social disparities in patient safety vary among social groups depending on the type of disease, treatment, or health service [9].

High-quality and safe care should be equally achievable for all patients and we need further studies to matching data on patient’s gender, educational, ethnic, and socioeconomical status with the data of critical incident registers.

1 Epidemiology of Adverse Event

Most published literature originated from the USA and the UK within the last 10 years.

The most common approaches for measuring harm in primary care include self-reporting by staff, analysis of existing databases, reviewing patient records manually or using automation, and asking professionals or patients to recall errors.

Most of these methods suffer from potential bias. Staff incident reports and patient and staff surveys are all affected by recall bias and potential social desirability bias.

Trigger tools are commonly used to identify events in hospital care but only a few studies have tested this approach in primary care [10, 11].

Furthermore, the majority of the studies published on medical errors in primary care show different error reporting methods and several definitions and classifications of types of medical errors (rate severity or preventability, etc.) and sources (estimated, legal reported or hospital referrals, etc.).

International reviews suggests that 1–2% of general practice consultations may result in adverse events, potentially preventable in 45–76% of cases, with estimated serious harm in 4–7% of all adverse event (resulted in permanent harm such as disability, death or long-lasting physical or mental consequences). Other authors report the occurrence of incidents to be between 5 and 80 times per 100,000 [3, 12,13,14].

In many studies, safety incidents are identified in three categories: administrative and communication incidents; diagnostic incidents; and prescribing and medication management incidents.

Some studies estimated that administration incidents occurred in at least 6% of patient contacts. Most of these incidents related to issues such as incomplete, unavailable, unclear, or incorrect documentation (coding/record keeping); inappropriate monitoring of laboratory tests (e.g., repeat blood tests for patients with repeat prescription or not checking results); or insufficient communication between professionals and patients (e.g. referral pathway for chronic diseases). Documentation errors are in high rate in undeveloped countries because of the lack of electronic records, while in developed countries they are less reported.

Diagnostic incidents are responsible for 4–45% of all patient safety related incidents. Common diagnostic incidents related to misdiagnosis or missed diagnoses and their effect can take months to years to manifest. Examining faulty clinical decision-making, shortcuts in reasoning (heuristics) emerges as an important entity. Misattributing presenting symptoms and signs to an obvious or readily available diagnosis may be a key issue, known as availability heuristics, or even anchoring heuristics, which occurs when doctors tend to maintain initial impressions once they are solidly formed. Any previous diagnostic label can reduce the clinician’s ability to reconsider an appropriate differential diagnosis list.

Researchers estimate an overall prescribing error rate from 3% to 65%, potentially occurred in any step of the medication process such as prescribing, transcribing, dispensing, administration, and monitoring (e.g., drug error, information error, or administrative error). Older persons are at higher risk than the general population for these adverse events and this probability increases to 75% in older people with four or more medications (polypharmacy) especially those residing in nursing homes. The reason of such high rate is probably related to elderly physiologic changes, frequent low health literacy, and drugs misuse for cognitive dysfunction.

2 Most Frequent Errors

The international medical community identifies three classes of factors as source of harm in primary care:

  • Human factors such as teamwork, communication, stress, and burnout.

  • Structural factors such as reporting systems, processes, and the environment.

  • Clinical factors such as medication.

The classification proposed here of the error areas is a function of the moment of citizen-professional meeting for simplicity here called “clinical moment” [15].

2.1 Preclinical Errors

Preclinical errors are due to the organizational activities of primary care professionals. In the preclinical setting, we can make further dimensional distinctions based on the specific competencies involved. We can thus distinguish access management errors of spontaneous citizens’ presentations (accesses postponed to days away for time-dependent pathologies with diagnostic and therapeutic delay; immediate accesses for self-resolving conditions with the risk of overmedicalization and overdiagnosis), administrative errors (errors of compilation and use of personal data sheets, error in the management and storage of sensitive data), reception problems (lack of or confusing reception and orientation to services both at the front office level and at the level of web information; lack or defect of architectural elements aimed at to accommodate the inconvenience of mobilization such as elevators, chairs, comfortable environments, absence of reception for people with audio-visual-linguistic barriers on a cultural or disability basis), problems linked to the physical achievement of primary care facilities (reachability by public transport or private, usability during day or night hours), problems related to the communication methods with the structure and the professionals (availability of telephone or web contacts, receipt of communications in the appropriate channels and quality of their management).

These dimensions at the risk of generating errors can variably cause damage in the short to medium or long term.

A citizen who fails to reach or communicate with the primary care facility may suffer a diagnostic or therapeutic delay, or turn to an improper setting with avoidable risks and improper use of resources to the detriment of other citizens.

A citizen who is not welcomed and oriented may delay the spontaneous presentation of the problem he perceives, may manifest anxiety or violent attitudes, or he can decide to postpone it and present it later with the risk of developing more demanding conditions.

This also has ethical, equity, and right to healthcare implications.

An incorrect management of personal data files, sensitive data, and poor care of electronic medical records can expose the patient to problems and damages even without a concurrent consultation with a professional. The risk of using erroneous or missing data can occur for exchanges in personal or administrative data or different clinical history used in subsequent meetings with professionals.

2.2 Clinical Errors

Clinical errors are those that are classically identified as a central element of safety problems, both among professionals and in the public. The authors agree that this type of error is an important part of the possible errors but not the most relevant.

Usually clinical errors in primary care have less impact than clinical errors in high-intensity settings; at the same time, the outcomes of these errors can reverberate at a distance.

A first group of errors concerns the prescriptive error of pharmacotherapy. LASA drugs (looks alike sounds alike) prescriptions are always possible and often avoided thanks to the electronic health record (EHR) software with its numerous checks necessary for the issue of the prescription and a last check at the time of delivery of the drugs in pharmacy. These errors often occur due to fatigue or distraction, rarely due to incompetence. To this end, EHR software, through its own alerts and security systems, can greatly limit errors with drugs and their dosages.

A second group of errors is the doctor’s diagnostic process. The errors of this group are in the scarce technical-medical competences during the medical examination and the anamnesis, the errors on the relationship and communication plan, the difficult management and communication of the diagnostic uncertainty, the erroneous prescription of lab tests and imaging tests. The latter represents a particular subgroup worthy of interest that calls for quaternary prevention. In the event that the primary care physician ask for a consultation to another doctor or prescribes further tests, this determines a certain risk of subjecting the patient to useless and sometimes harmful medical practices: unnecessary examinations, increased costs and time, over-use of resources of the health system and patient, sensitive iatrogenic risk investigations such as those with ionizing radiation, and accesses in specialized healthcare settings.

A third group of errors is the malfunctioning of the team or its ineffective use and its lack. It is internationally shared options that primary care established in multidisciplinary teams where more skills can be present at the patient’s closest level may be able to respond to health needs. Malfunctioning teams with staff shortages or lack of leadership can generate clinical errors: lack of sharing and planning of care plans, failure to share information, non-assignment and recognition of professional responsibilities.

A fourth group of errors consists of delayed diagnoses. As said for the second group the diagnosis is a social fact, an error placed in its social determinants will produce a clinical error. An example is precisely delayed diagnosis or delayed prevention. Both underlie the presence of a lack of clinical and organizational competence. By way of example, the diagnosis of advanced diabetic foot lesions can be characterized by a diagnostic delay in initial lesions or even by a failure to prevent the diabetic foot (an activity that must be organized) if not precisely by the failure to prevent diabetes. It appears obvious here that the question of error goes beyond the ideal boundaries of primary care as the near or distant responsibilities lie in individual behavior and in the social determinants of disease, but it is equally evident that the problem emerges in the clinical setting of primary care among physicians and patient.

A fifth group of errors is in the promotion-preventive sector. The banal absence of preventive and health-promoting activities is notorious that involves development of disease and represents a mistake. This type of error is intertwined in preventive skills and competences and in organizational skills. This is a very important point because in the medicine of the twenty-first century the attention and resources to prevention must grow in order to pass effectively from an acute organization to an organization for complex problems and chronic cases.

3 Clinical Cases

3.1 Clinical Case: Being Alert

The wife of a 68-year-old patient calls the switchboard of the group medicine because her husband has been suffering from dyspnea for a few days and is worried, he calls at 4.00 pm and the doctor invites him to present himself the same evening, having time to dedicate to him. The patient arrives after about an hour, on foot (his house is about 1500 m away) accompanied by his wife and daughter. It appears dyspnoic. The patient is known to the doctor, has a low socioeconomic condition. As a chronic diseases, he has type 2 diabetes mellitus in oral antidiabetic therapy, COPD 2b in inhalation with LABA and LAMA, he has an essential tremor on propranolol therapy. The doctor investigates any concurrent events, the patient denies fever denies trauma and denies coughing. He denies chest pain. The vital parameters are good (PA 130/70 FC 75r Sat 97% T 36.3). Being a patient with an elevated cardiovascular risk (previous tobacco smoker, diabetic) the doctor proceeds with the execution of an ECG in the office. The ECG demonstrates an elevation of V1, V2, V3, and V4 and suspecting an acute coronary syndrome alerting the territorial emergency-urgency service for rapid access to the Emergency Room. Furtherly the doctor will be informed by the children about the diagnostic confirmation, which will be followed by an urgent coronary angiography with medicated DES and then the setting up of an antiplatelet therapy and cardiological rehabilitation.

Compared to the safety practices adopted we can identify some hidden but absolutely important elements.

The availability for urgent visits is the essential element that allows citizens to be able to have a confrontation with the doctor for situations that alarm them but that do not lead them to independently access emergency services. This availability cannot be causal but is the result of an organization, and therefore of an organizational, multidisciplinary competence that involves physician and ancillary staff (front office). The possibility of receiving phone calls during daytime means setting up a service capable not only of receiving phone calls during daylight hours but also of making the doctor interact with these requests for attention. This service is substantiated with the presence of dedicated staff and doctors available during the daytime hours. A triage, run by medical or nursing professionals, can be added to this organization, which could significantly increase the quality of the management of urgent requests, but today there is little literature on the subject and in any case does not fit into the clinical case scenario.

The presence of time and space is an element of safety. The organization of primary care must foresee the unforeseen and equip itself with effective response capacities: it is necessary that unpredictable visits can take place in times and places consistent with the need for each single clinical case.

On the instrumental level we can see how a doctor experienced in the diagnostic use of medical equipment represents a critical element. For the patient’s condition, probably the doctor would still have had to send it to the emergency service as it would have been considered necessary the troponin curve not feasible in another setting. But the early diagnostic suspicion made it possible to activate the further medical service in order to optimize the management with times and strategies typical for a life-threatening condition. This technical element, of medical competence, conceals the element of safety that concerns the continuity of care in setting transitions which also represents a competence, though not merely clinical.

3.2 Clinical Case: A Foreseeable Error

In the hottest days of summer, Aldo, a diabetic 83-year-old man, is recovering from a weekend gastroenteritis with 2 days of vomiting and diarrhea. He is not worried because his nephew had a similar condition few days before and told him what to do. In a certain way, he hopes to lose weight faster and reduce his high glycemic level. He didn’t take all medications in the last days, except those for blood pressure and diabetes (ACE-I and metformin). He tried to call his doctor, but the line was always busy and the secretary told him that first office appointment available would have been the next week, so he decided to postpone the consultation, also because he is tired and his back pain has increased in the past few days. Before going to bed, he takes three pills of an over-the-counter painkiller (NSAID) to reduce back pain. Two days after, his fatigue has increased and he urinates very little, almost nothing from midnight, even if he has doubled the dose of diuretic. His worried nephew tells him call the doctor but Aldo decides to wait until the afternoon, when the doctor is in office, but he is a little confused and go to bed and he sleeps all evening waking up only for dinner. Aldo is very tired and little inclined to speak and also he doesn’t trust the young doctors on call. When he calls, he only talks about his tiredness The doctor who answers doesn’t investigate about his comorbidities or medication and reassures him saying that this is the normal course of acute viral gastroenteritis.

During the night, Aldo is not doing well, he is confused, extremely tired, sick, and when dyspnea occurs, he decide to call the Emergency Department. Admitted in the Emergency Room, his diagnosis is acute kidney injury.

Acute kidney injury (AKI) is a sudden episode of kidney failure or kidney damage that happens within a few hours or a few days. AKI occurs in approximately 10–15% of patients admitted to hospital. It is a serious condition with fourfold increased hospital mortality. It is defined by a rapid increase in serum creatinine, decrease in urine output, or both. Comorbid conditions including CKD, diabetes, hypertension, coronary artery disease, heart failure, liver disease, and chronic obstructive pulmonary disease are risk factors for AKI as well as age (over 65 years), nephrotoxic exposures (iodine products and drugs as NSAIDs, ACE-I, diuretics, etc.) major surgery, sepsis, fluid resuscitation, and volume status (dehydration by Inadequate fluid intake, excessive vomiting, diarrhea, and fever).

In this clinical case, Aldo had not been sufficiently informed of the side effects of the medicines he was taking or of the measures to be taken in the event of a fever or vomiting or diarrhea. Prescribing a medicine, with reviewing side effects, contraindications, and drugs interactions, is an essential part of the primary care visit. Using a tool to monitor that step during the visit would be useful to avoid medication incident. Another communicative incident is also the difficulty of accessing a medical consultation with his primary doctor, both by phone and by visit. The lack of confidence in the unknown doctor leads him to silence parts of his condition and the doctor on call does not sufficiently investigate patient’s condition (age, comorbidities, etc.). This diagnostic error is due to both clinical and organizational competence. The incomplete anamnestic review leads to superficial diagnosis and underestimated severity with potential serious harm (from intensive care to dialysis or death). A more organized primary care team would have shared medical records, avoiding misunderstanding.

A further element of safety that those scenario predicts, but does not described, is the possibility of a review of the case through the SEA model (significant event analysis). The diagnosis, during the acute phase, of a life-threatening disease is a significant event in primary care. This phenomenon represents an excellent opportunity for the multidisciplinary team to review its skills and competences and understand if mistakes have been made, regardless of the positive outcome of the specific case.

4 Safety Procedures

Safety is a major concern in four main areas: diagnosis, prescribing, communication, and organizational change [16].

4.1 Diagnosis

Diagnosis in primary care is by its very nature uncertain and uses a hypothetic-deductive approach. A general practitioners deals with a very broad range of symptoms and signs with no clear diagnosis in most cases and his longitudinal care leads to practice related to individualized needs, preferences, and values. Therefore, the use of guidelines and protocols is likely to have some, but limited, success in improving safety. Some procedures are known to be safer than others, and those could be the best practices to improve in primary care. Decision support tools and (electronic) information systems can be hypothetically useful, but this has not yet been proved empirically. But, many safety problems can be overcome by design, for example, the use in specific circumstances of reminders such as message alerts on screen or insertion of checks or forcing steps instead of relying on memory and observance.

4.2 Prescribing

Prescribing is the most analyzed area. Hospital-based studies have shown that use of a computer system for prescribing is likely to improve accuracy. In many computer medical records, there is the opportunity to highlight possible drug-drug interactions, individual known drug hypersensitivities, and relative and absolute contraindications related to clinical conditions. But, many computer systems currently use alerts so often that many doctors simply choose to ignore them (the “cry wolf” phenomenon) and the increasing use of complementary treatments including herbal remedies is often not reported by patient because they are not considered medicine.

4.3 Communication

Communication is a common cause of harm to patients, but it is probably a symptom of organizational problem rather than a cause. Medical errors can occur due to a lack of communication both between colleagues and between doctor and patient.

Electronic communication can reduce problems in sharing clinical information or therapies or allergies among clinicians. The “patient held record” (better if held on the internet) would ensure that clinicians have immediate access to all relevant clinical information and assure consistency across primary and secondary care.

A poor doctor–patient relationship can have negative outcomes for patient satisfaction, treatment compliance, and even the health status of the patient (missed or inadequate diagnosis, for example). Agreed methods of communicating would be developed and established at the first visit, well known by team members and secretarial staff. Each therapy should be reviewed and re-explained at each visit in writing form. While respecting patient privacy, all clinical information should be known to the primary care team.

4.4 Organizational Change

The reporting of incidents can help healthcare professionals learn from mistakes. Leaders within the system should reward and encourage doctors to report problems in order to take specific action to prevent the problem occurring again. The understanding how system fails, reporting and analysis of the medical errors, the use of technology, and the continuous attention to the safety culture can lead to a quick improvement of the primary care. Where primary care is organized in teamwork, it has the ability of sharing and analyzing medical errors, actual or perceived, and implements those organizational changes necessary for a better development of the safety culture [17].

5 Recommendations

Patient safety incidents (PSIs) in primary care are perceived as a relatively lower-risk endeavor, but about 4–7% of the errors have the potential to cause serious harm, either in the short or long term. Most of them are preventable. In secondary care settings, targeted strategies have been implemented and reported but in primary care we have limited data and the World Health Organization has noted the pressing need to study and address patient safety in this setting.

As Dr. De Wet pointed out, improving patient safety in general practice requires “action on at least three fronts: greater evidence-based knowledge of patient safety, time and space to conduct the required, appropriate reflection and a strong safety culture within practices, characterised by excellent leadership, effective communication, and team members who support each other and learn together” [14].

The importance of “human factors” and the complexity of doctor–patient interactions in primary care can influence any health system and need to be investigated to better understanding their role in causing patient safety incidents.

To reduce harm and improve patient safety is essential to overcome the diversity in the reported frequency and nature of errors and to develop an understanding of both the causation and prevention of error in primary care. Actually, we have an opportunistic incident reporting rather than a systematic and proactive approach and we need a more specific intervention in the definition of “error” in primary care, a common rate for severity and preventability and in collecting data.

In 2019, González-Formoso shows that education is a key pillar of quality improvement and is considered the most important factor in improving patient safety, especially in primary care where the effectiveness of the educational intervention given to residents and their tutors in family medicine teaching units was measured by the number of events reported [18]. Training in patient safety improves knowledge and the process of care. The effectiveness of specific interventions to reliably reduce harm in general practice remains unknown. Further studies are needed to examine whether and how the professionals participating in the educational intervention have modified their behavior with respect to patient safety and whether patients’ outcomes did improve.