Autoimmune haemolytic anaemia due to immunodeficiency

Abstract

Autoimmune disorders are common presenting manifestations of immunodeficiency syndromes. We present a case of a woman in her late teens, with a history of frequent sinopulmonary tract infections during her childhood, who presented to our hospital with anaemia, jaundice and fatigue. She also had significant physical growth retardation for her age and sex. With this case report, we intend to present the diagnostic and therapeutic challenges faced by the patient and our healthcare system and propose a few feasible solutions to tackle these challenges.

Background

Primary immunodeficiency disorders (PIDs) are a common but significantly underdiagnosed and under-reported group of heterogeneous disorders.¹ Constituting this group are 176 conservatively defined hereditary disorders,² which affect the immune system or its functions.³ Common variable immunodeficiency (CVID) is a group of such PIDs, characterised by a failure in B-cell differentiation, with a common phenotype of sinopulmonary tract infections.⁴ Several studies have also shown an increased predisposition to mycoplasma infections and its association with autoimmunity in patients with CVID.^5–7

In 2013, the WHO published an analysis,⁸ which estimated the frequency of PIDs in various countries and continents—they suggested that there may be 6 million people living with a PID worldwide, of which only 27 000–60 000 were identified and publicly reported.

Coupled to this, are the various non-infectious phenotypes of PIDs—such as allergies, autoinflammation, autoimmunity, cancer and several such diverse phenotypes.⁹ Thus, the diagnosis of PIDs requires a thorough understanding of disease epidemiology and clinical characteristics.

A multicentre prospective study by Quinti et al¹⁰ reported the natural history of CVID in 224 patients. They observed that autoimmune manifestations were found in 17.4% (39/224) of these patients before a diagnosis of CVID was made. Sicca syndrome, autoimmune thrombocytopenia (ITP) and autoimmune haemolytic anaemia (AIHA) were the three most common autoimmune manifestations, in that order. After diagnosis and during long-term follow-up (mean time—11 years), a further 19 patients (total 58/224–25.9%) developed autoimmunity. Twenty-one out of 58 patients (36.2%) were younger than age 14 years.

Possible pathophysiological mechanisms include—abnormal B cell maturation and defective development of class-switched memory B cells and the development of CD21^low B cells.^11–16 The absence or low numbers of these subpopulations of B cells have been associated with splenomegaly, autoimmune cytopenias, granulomatous disease and lymphadenopathy.¹⁷ Although the exact pathophysiological mechanism remains unclear, one study¹⁸ showed that CD21^low B cells predominantly expressed autoreactive antibodies to nuclear and cytoplasmic structures.

Hence, a clinical history of frequent sinopulmonary tract infections and autoimmunity in young patients should elicit suspicion of CVID or a PID. The natural history of CVID merits long-term follow-up, usually over several years, high-quality and precise documentation of infectious and non-infectious manifestations and subsequent early detection and prompt institution of therapy for system-specific manifestations (rheumatological—vasculitis, systemic lupus erythematosus, rheumatoid arthritis (RA), etc; gastrointestinal—inflammatory bowel disease, pernicious anaemia, autoimmune hepatitis, etc)¹⁹

We present a case of a woman in her late teens, with anaemia, jaundice and profound fatigue. A diagnosis of AIHA was made. To further ascertain the cause of AIHA, we had to go through a large volume of poorly and imprecisely documented paper-based medical notes, to make a final diagnosis. This paper aims to highlight the difficulties faced by the patient throughout her medical journey and the difficulties faced by the healthcare system in maintaining data continuity and in arriving at a final diagnosis.

Case presentation

A young woman in her late teens presented to our hospital with 3 days of exertional dyspnoea, yellow discolouration of the eyes and severe fatigue. On further questioning, she also reported palpitations, early satiety and bendopnea (breathlessness on leaning forward). She denied signs and symptoms of cola-coloured urine, fever, acute loss of appetite or weight loss. She also denied pruritus (itching) or vomiting.

The patient’s clinical history begins when she was a toddler when she first started developing symptoms of upper respiratory tract infections (rhinorrhoea, sneezing, cold and headaches). At around the same time, she also started developing frequent fevers, and cough with copious expectoration and was diagnosed with acute bronchitis several times.

One year later, the patient developed a fever, cough and poor appetite. She was diagnosed with a right middle lobe consolidation and was treated with antibiotics and other supportive therapy for fever, cough and loss of appetite. No records of subsequent follow-up were available.

A few months later, the patient developed periodic fevers with sweating. She was diagnosed with smear-positive falciparum malaria and was treated for it.

One year later, she was diagnosed with chronic suppurative otitis media and was treated for it. No records on the causative organism and subsequent follow-up were available. Shortly after, she was diagnosed with a lower respiratory tract infection (LRTI) and was treated with antibiotics and other supportive therapy for fever, cough and malaise.

One year later, she was diagnosed with bilateral bronchopneumonia and was treated with cefpodoxime syrup. A few months later, she again developed a fever and cough with poor appetite. A Mantoux test was done, which showed a 12 mm induration after 48 hours. She was presumptively diagnosed with pulmonary tuberculosis (TB) and was administered antitubercular therapy.

Over the next few years, the patient had several such episodes of recurring upper and LRTIs. A few years later, she was diagnosed with scabies and was treated for it.

At around the same age, when she had another LRTI, the treating physicians noticed her failure to thrive with severe stunting and wasting for age and sex.

She continued to have frequent sinopulmonary tract infections, with regular visits to hospitals. At that time, she weighed 11 kg and 2 years later, she weighed 13 kg.

A few years later, the patient again reported a mild fever and cough, with scanty expectoration. At this point, she reported jaundice for the first time in her life. A review of her paper-based medical notes revealed that she was provisionally diagnosed with viral hepatitis and was treated for it. She also had severe symptomatic anaemia during that admission and received one packet of packed red cells transfusion. After discharge, she had frequent respiratory tract infections requiring several visits to multiple hospitals.

Four years before the current presentation, she presented to our hospital with a similar history of fever, cough and expectoration. She was diagnosed with a suspected LRTI and was treated with antibiotics.

Over the next 2 years, she continued to suffer from similar symptoms of lower respiratory tract involvement. She was treated at multiple different centres, which only provided temporary relief, according to her.

A few months later, the patient started developing jaundice and profound fatigue with breathlessness. Tests were performed and the patient was diagnosed with AIHA.

Initially, all presentations to medical facilities were within the same geographical area, most of which were government healthcare facilities. However, since the development of anaemia, she visited private healthcare facilities in the city close to her native geographical region. Notes from these places show that data continuity was not maintained and there was also no evidence of acknowledgement of past medical data in the notes from these healthcare facilities. Although all medical data were recorded in English, patient counselling was always done in her native language. Throughout these visits, the patient never had a single lead clinician with whom she could have followed up regularly. However, since her presentation to our hospital, our team became her primary clinician team, and all subsequent medical consultations and data logging was performed by the same team. A detailed summary of her clinical history and subsequent clinical events is outlined in table 1.

Table 1.

Summary of clinical events

Event no	Stage of life	Clinical event	Phenotype	Follow-up	Detection of failure to thrive
1	Toddler	Upper respiratory tract (URT) symptoms	Infectious (presumed)	×	×
2	Toddler	Lower respiratory tract (LRT) symptoms	Infectious (presumed)	×	×
3	Early childhood	Right middle lobe consolidation	Infectious (presumed)	×	×
4	Early childhood	Malaria	Infectious	×	×
5	Early childhood	Chronic suppurative otitis media	Infectious	×	×
6	Early childhood	LRT symptoms	Infectious (presumed)	×	×
7	Early childhood	Bilateral bronchopneumonia	Infectious (presumed)	×	×
8	Early childhood	Pulmonary tuberculosis	Infectious (presumed)	×	×
9	Early-middle childhood	Multiple episodes of URT and LRT symptoms	Infectious (presumed)	×	×
10	Middle childhood	Scabies	Infectious	×	×
11	Middle childhood	LRT symptoms	Infectious (presumed)	×	✔
12	Middle childhood	Multiple episodes of URT and LRT symptoms	Infectious (presumed)	×	✔
13	Early adolescence	Anaemia, jaundice and fatigue requiring one blood transfusion. Provisionally diagnosed as viral hepatitis	Infectious (presumed)	×	×
14*	Early adolescence	LRT symptoms. Treated with antibiotics. A possibility of allergic bronchitis is considered for the first time	Infectious (presumed) & non-infectious	✔	×
15*	Early adolescence	Multiple episodes of persistent LRT symptoms	Infectious & non-infectious	✔	×
16	Early adolescence	Anaemia, jaundice and breathlessness—first diagnosis autoimmune haemolytic anaemia (AIHA)	Non-infectious	✔	×
17	Early adolescence	Attains menarche. Irregular menstrual cycles +	–	–	–
18	Early adolescence	Anaemia and jaundice—AIHA	Non-infectious	✔	×
19	Early adolescence	Anaemia and jaundice? Idiopathic AIHA. Required two blood transfusions	Non-infectious	✔	×
20	Early adolescence	AIHA. Referred to gastroenterology for further workup. Upper gastrointestinal (GI) endoscopy—normal. Haemoglobin (Hb) electrophoresis—normal	Non-infectious	✔	×
21	Late adolescence	Multiple consults at various hospitals for anaemia. B12 and folic acid levels, osmotic fragility test, Hb electrophoresis (repeated), whole genome sequencing and a battery of other tests repeated	Non-infectious	✔	×
22*	Late adolescence	Anaemia and jaundice – AIHA. Presented to us.	Non-infectious	✔	✔

*Presentations to our hospital. Events 14 and 15 were at the department of pulmonary medicine and event 22 was at the department of general medicine. Although a diagnosis of CVID was not suspected in her earlier presentations to our hospital, data from those presentations enabled us to consider a diagnosis of CVID. No direct communication occurred between the clinicians of these two departments. Her immunisation status was unknown.

CVID, common variable immunodeficiency.

At admission, the patient reported being comfortable at rest, with mild respiratory distress. She appeared pale, stunted and wasted for age and sex. Her vitals were—a pulse rate of 123 beats per minute (bpm), blood pressure—of 110/50 mm Hg, respiratory rate—of 34 breaths per minute and a temperature of 98.4° Fahrenheit (37.2°C). She had a room air oxygen saturation of 97%. She was 140 cm tall and weighed 25 kg (figure 1).

Figure 1.

Growth charts for height and weight for our patient. Our patient was significantly stunted and wasted, with both height and weight below 2 SD of mean. The aetiological factor for these was likely recurrent and chronic childhood infections.

General physical examination revealed conjunctival pallor, scleral icterus and bilateral grade 2 pitting type pedal oedema. Examination of the cardiovascular system was significant for raised jugular venous pressure, a loud P2 (pulmonary component of the second heart sound) and a grade 3 pansystolic murmur in the mitral and tricuspid areas. She also had a grade 2 parasternal lift and palpable pulsations in the epigastric regions. Abdominal examination was significant for palpable splenomegaly, 3 cm below the costal margin in the anterior axillary line. Examination of the respiratory system revealed reduced chest expansion on the right side with decreased air entry in the right infrascapular area. Neurological examination was normal.

After all her paper documents (which she had with her) were reviewed and the history and examination at current admission, a provisional diagnosis of CVID was made. The AIHA was attributed to an autoimmune manifestation of the PID.

Her haemoglobin (Hb) was at 32 g/L; normal range=110–150 g/L and the total leucocyte count was at 8200 cells/mm³(8.2 x10ˆ9 cells/L). Her platelet count was normal (normal range=150 - 450 x10ˆ9 cells/L) and red cell count was 0.8 x10ˆ12 cells/L(4.5 - 5.5 x10ˆ12 cells/L), with a mean corpuscular volume of 116 fL (normal range=80–100 fL). The red cell distribution width was elevated at 25 (normal=11–15). The packed cell volume (PCV) was 9%. Reticulocyte count was elevated at 8%.

Her total bilirubin was 8.6 mg/dL, with a direct bilirubin of 0.77 mg/dL. Liver enzymes were within the normal range and a narrow gamma gap (serum total proteins minus serum albumin levels) of 1.3 was noted, with the total proteins at 5.9 g/dL and albumin of 4.6 g/dL (normal range of gamma gap=2.5–3.5).

Peripheral blood smear showed macrocytes, macro-ovalocytes and anisopoikilocytosis with hypochromia. A direct Coomb’s test was performed, which showed 4+ reactivity (figure 2).

Figure 2.

Direct Coomb’s test—circled test tube shows 4+ agglutination indicating presence of autoimmune antibodies against red cell self-antigens. Rectangle marked test tube shows no evidence of haemagglutination. Triangle marked test tube shows 2+ haemagglutination.

A chest X-ray was done, which showed tiny hyperlucent areas with tram tracking in the right lower zone, in the periphery, suggestive of bronchiectasis, likely secondary to recurrent lower respiratory tract inflammation. A two-dimensional-echo confirmed a high output heart failure secondary to severe anaemia—the right atrium, right ventricle and left atrium were dilated, and the ejection fraction was 72% (figure 3).

Figure 3.

Chest X-rays posteroanterior view (left) and right lateral view (right)—the X-rays show bilateral perihilar hyperlucent ring shadows and cystic dilatation of bronchus in the right lower zone. The right lateral view demonstrates dilated right lower lobe bronchus visible upto <1 cm from the pleural marking. These features are consistent with bilateral cystic bronchiectasis likely due to recurrent childhood infections.

In view of severe symptomatic anaemia with profound fatigue, a blood transfusion was planned, and dexamethasone injections and other supportive therapy (furosemide for relieving fluid overload due to heart failure) were given in the interim. Several blood samples had to be tested because the patient’s blood was cross-reactive to most. One unit with the least cross-reactivity was found and transfused uneventfully. Post-transfusion, the patient’s symptoms resolved, with her Hb improving to 56 g/L and a PCV of 15%.

X-rays of the hands were taken to check for bone growth and age. Her bone growth was consistent with those of normal 17–21 years old people (figure 4).

Figure 4.

Bilateral hand X-rays taken in frontal views show diffuse increase in bone density in the proximal and middle phalanges of all fingers. Bone age based on bilateral hand X-rays—Distal epiphyses of both radius and ulna are fused suggesting a bone age of 17–21 years. Metacarpal epiphyses, proximal and distal phalanges closure have occurred, suggesting a bone age between 14 and 21 years.

For a diagnosis of CVID, both serum complement and immunoglobulin levels were checked. C3 and C4 complement levels were normal but immunoglobulins G, A and M (IgG, IgA and IgM) were all decreased. Her serum IgG levels were 3 g/L (normal range=7–16 g/L); serum IgA levels were<0.1 g/L (normal range=0.7–4 g/L) and serum IgM levels were <0.2 g/L (normal range=0.4–2.3 g/L).

For a diagnosis of CVID, the definition published by the European Society of Immunodeficiencies (ESID) and the Pan American Group for Immune deficiencies (PAGID) in 1999 is commonly used.²⁰ Ameratunga et al²¹ presented a critical analysis of these criteria and highlighted a few pitfalls:

• IgG levels less than 2 SD below the mean can lead to overdiagnosis as 2.5% of the general population would have this.

• No clear definition of what an impaired vaccine response is, and which type of vaccines generate a poor immune response.

• Exclusion of secondary causes of immunodeficiency—secondary causes of immunodeficiency can be numerous and excluding secondary causes cause a strain on both the patient and the healthcare system.

In response, the authors proposed new diagnostic criteria involving multiple domains (three major criteria, clinical sequelae directly as a failure of the immune system, supporting laboratory evidence and histological markers of CVID; the last domain of histological markers is not required for diagnosis; it only increases the probability of CVID). Our patient met all the three major criteria and clinical and laboratory criteria required for diagnosis.

At discharge, the patient’s symptoms had resolved, her Hb improved to 89 g/L (8.9 g/dL) and her jaundice had resolved. The patient and her attendants were counselled about intravenous Ig therapy. The patient was asked to remain under follow-up with us and an electronic medical record (with due consent) where notes from all her paper records were converted to an electronic patient record, with a clear sequence of events, was created and shared with the patient.

Global health problem list

1. An urgent need for the introduction of centrally accessible electronic medical records.

2. The questionable efficacy and evidence base for recommending immunoglobulin replacement therapy (IRT).

3. Precision diagnosis in India.

Global health problem analysis

An urgent need for the introduction of centrally accessible electronic medical records

‘Listen to your patients—they’re telling you the diagnosis!’—William Osler famously said this to his students in the late 19th century.

Osler was perhaps the first physician to understand the importance of the natural history of the disease.²² An analysis of table 1 presented above lays bare data discontinuity and the importance of a well-kept medical record, which provides access to a patient’s medical consultations, sequentially.

Coupled with this, India, being the second most populous country in the world, has a disturbingly unequal distribution of healthcare facilities across various sectors. About 75% of all our healthcare infrastructure is in urban areas, where only 27% of the population resides.²³ Thus, although India has taken progressive measures in improving access to healthcare, a disparity in its distribution is making access to quality healthcare difficult for rural populations.

Thus, it becomes imperative that healthcare workers (HCWs) prioritise cost-effectiveness and easier and better access to quality healthcare. Das et al²⁴ showed that although ‘informal’ HCWs (‘self-declared’ doctors in India—without any formal medical education but with alternative routes of medical certification) provided inferior quality care when compared with formal medical practitioners, a significant improvement in case management and logging of patient data was demonstrated. A few other notable studies have also demonstrated the utility and effectiveness of training rural HCWs in logging patient data on online platforms and subsequent provision of these data to trained, formal HCWs for analysis and directing appropriate care.^25–27

Thus, the true principles of Oslerian medicine can be followed—patient-related events are recorded and presented longitudinally and the natural history of the disease and its impact on the patient’s life can be better understood.

The National Health Policy (India) 2017 highlighted this issue and sought to address this with the introduction of a health ID for any and every patient in India. This health ID (a.k.a. Ayushman Bharat Digital Mission) ‘will be used to uniquely identify persons, authenticate them and thread their health records (only with the informed consent of the patient) across multiple systems and stakeholders.’²⁸

Although this ostensibly ambitious plan was well-received across the country,²⁹ several potential barriers need to be addressed, some of which are:

• No proof of concept on the design and operation of the system—this programme attempts to transform the current system of subcentres, primary healthcare (PHCs) and community healthcare systems into Ayushman Bharat health and wellness centres (comprehensive PHC centres delivering a range of services spanning preventive, promotive, curative, rehabilitative and palliative care).²⁸ Data from the north Indian state of Punjab show that—infrastructure, training of healthcare personnel, community participation and state and local authorities’ involvement—are potential barriers to successful implementation.³⁰

• A commitment to use blockchain technology (recording data in ‘blocks’ sequentially from the original node to the destination node. These blocks are linked together securely preventing any block from being altered or a new block being inserted between two existing blocks), which is not in widespread use in healthcare industries yet.³¹

• The project aims to integrate modern medicine with alternative forms of medicine (such as AYUSH—Ayurveda, Unani medicine, Siddha medicine and homoeopathy). A clear roadmap on how this can be achieved has not been laid out yet. Although this patient never sought alternative forms of medicine, it is common practice here in India to seek alternative forms of medical care.³² Paper-based records are sometimes kept and clinical findings are rarely documented and only the medicine prescribed is documented.³³

Thus, one of the key aspects of this programme—to implement electronic health records (EHRs)—would enable longitudinal data documentation and improve precision in diagnosis.³⁴ This also enables asynchronous communication, a team of physicians taking the role of primary care clinicians for a patient and multiple caregivers documenting clinical notes.³⁴ On the other hand, EHRs bring more transparency in documenting diagnostic errors and auditing performance.

A feasible solution we have employed with measurable success and would like to suggest to the wider community is the use of easy, widely established, publicly available and open access third-party platforms (such as blogger blogs) to document patient data in a deidentified manner after due consent is taken from them. We identified several advantages with these, which are laid out below:

• Requires minimal training in the use of technology and know-how to document data.

• Follow-up data can be documented easily.

• Patients themselves can view their records and the burden of carrying thick bundles of paperwork is taken off.

• Patients themselves can create their own medical records and edit them as and when needed.

• Academic discussions and the rationale for decisions can be added directly to the record, maintaining continuity of data logging.

• An informal portfolio, to assess a student’s core clinical competencies.

Some potential pitfalls, however, exist:

• Proper deidentification of data is paramount.

• Documenting sensitive data on third-party platforms can cause privacy issues.

• Requires teamwork and all the doctors involved in patient care to embrace this system (figures 5 and 6).

Figure 5.

Image of data documentation using a blogger blog. Clinical data of the patient can be documented by a team of caregivers and can be accessible by other caregivers and the patient too. Image created with permission from primary author of patient’s blog.

Figure 6.

Similar image of clinical data documented on a blogger blog. Information continuity is demonstrated through the use of charts. Image created with permission from primary author of patient’s blog.

A few other studies have also reported on the potential pitfalls of large-scale implementation of EHRs—data breaches, doctor burnout and interoperability between various EHR software are some stumbling blocks.³⁵

One study from Karnataka, India, attempted the implementation of EHRs to monitor maternal and child health (MCH), and to track process indicators in MCH.³⁶ EHRs were compared with paper-based records in the recording data on these process indicators (total count of 12 processes). HCWs had to document these on both the app and paper-record. The study was done for 2 years, and the results showed that recording of data on paper was easier and preferred by most HCWs (answered in interviews) and that the dual burden of recording data electronically and on paper was one of the most common reasons for not adopting EHR consistently. Only 3/12 process indicators were recorded with more than 80% consistency when compared with paper records. Recording of data on stillbirths, neonatal deaths and mothers who received postnatal care showed the lowest consistency. The study also reported that government audit of paper-based records was the primary reason for preferring paper-based records.

A report, published by the ministry of Electronics & Information Technology, Government of India,³⁷ laid out a roadmap for successful and wider adoption of EHRs. The report suggested that inter-operability between EHR software, protection of privacy, open sourcing EHR software and allowing patients or citizens to maintain a personal health record system, themselves should be prioritised for wider adoption.

While EHRs may improve data documentation and subsequent utilisation of that data for future audits and analysis, the larger question of whether EHRs improve the quality of healthcare remains unanswered.

One study from Jordan³⁸ recruited 410 healthcare professionals, 205 each from two different hospitals. Sixty-three per cent of those 410 were nurses and about 31% were physicians. They compared EHRs against paper-based records in the following domains—tangibility (ease of use), reliability, responsiveness, assurance and empathy, using Likert scales (with 1 being lowest and 5 being highest). The results showed that expectations and perception of improved quality of services were negative for both, with a mean difference of −0.13 among all domains in the EHR group vs −0.34 among all domains in the paper-based group. In both groups, responsiveness (reducing time to access healthcare services) and empathy showed the highest differences between expectation and perception.

While the above study compared the quality of healthcare services on subjective scales, one Canadian study³⁹ compared EHRs with paper-based records in improving the delivery of certain healthcare services, over 2 years (2006 and 2007)—Papanicolaou smear, screening mammograms, influenza vaccination, faecal occult blood test and five primary immunisations for children. Baseline characteristics of physicians embracing EHRs versus paper-based records were similar among both groups (such as year of graduation, number of hours worked per week, number of patients per physician etc). The study recruited more than 34 000 patients and results showed that there was a negative difference in administering influenza vaccines from 70.7% in 2006 to 69.8% in 2007 when compared with the non-EHR group (59.2% in 2006 to 66.5% in 2007), and a similar trend was also observed in administering a faecal occult blood test. A marginal overall increase of +1.1% was noted in administering Papanicolaou stain, however, a significant difference of+6.3% was noted in administering mammograms.

A few other authors^40–42 have suggested that implementation costs, software immaturity, lack of connectivity with external systems and lesser freedom to document medical notes to one’s style as some reasons why EHRs may not necessarily improve the delivery of healthcare.

Another useful aspect of EHRs is the generation of easily accessible data for audit and analysis. Baron’s study⁴³ showed an improved mammography rate by 10% with the implementation of EHRs, when compared with paper-based records. The author, however, notes that they had to buy expensive software to scan and convert those paper-based data to electronically accessible ones. He also suggested that some data are much more difficult to scan and convert (such as funduscopy findings) than other data. In their study, their initial estimation was that the mammogram rate of women in the age group of 50–65 years, presenting to their hospital was 50%. However, once all data were organised and structured to be read by computers, they noted that their actual mammography rate was 65%. Automated reminders programmed by the EHRs improved the mammography rate to 75%—an absolute increase of 10%.

Thus, with time and training, data accrued and structured for accessibility by computers can help achieve quality improvement.

EHRs, while not the one-stop solution for all healthcare delivery problems, can improve data transparency, and accountability⁴⁴ and generate vital data for audit and quality improvement. However, it brings several other issues such as the requirement of lengthy periods of training for operability, lack of connectivity with external systems, no cross-compatibility with other systems and possibly reduced responsiveness and empathy from doctors.

Whether EHRs improve the quality of healthcare is still debatable, the generation of centrally accessible data should enable periodic audits and subsequent improvement. In our case, a blogger blog was created, and all patient-related medical data were logged, this enabled a longitudinal approach which clinched the diagnosis and data from subsequent visits too was logged. However, as noted above, privacy issues always remain a concern, data are only logged at the time of visit to our hospital, data continuity may be lost with a visit to another hospital and the process of electronic documentation, apart from maintaining paper-based records is a laborious process.

The questionable efficacy and evidence base for recommending IRT

The ESID/PAGID guidelines recommend the use of IRT in patients diagnosed with CVID.²⁰ The evidence base on which this recommendation is made is, at best, questionable. No single, placebo-controlled randomised trial exists, for making a strong recommendation for IRT use.

Several observational studies have been performed to date^45–48 with no uniformity in methodology or targets for hard clinical endpoints. Quinti et al⁴⁹ performed a prospective observational study on 201 patients diagnosed with CVID according to ESID/PAGID criteria. Patients were longitudinally followed up for at least 5 years. The mean age at diagnosis was 28.7±18.4 years. The dose of IRT varied from 400 mg/kg/month to 1000 mg/kg/month. In some patients, an IgG trough level of >1000 mg/dL was targeted. Observed endpoints included incidence of pneumonia, acute and chronic sinusitis, bronchiectasis and any invasive bacterial infections. Overall, 79% of patients had major infectious complications (at least one episode of the above-mentioned endpoints). A significant reduction was noted in the incidence of invasive bacterial infections, after initiation of IRT (a risk reduction from 16.6% (preinitiation of IRT) to 2.5% (postinitiation of IRT)). The incidence and prevalence of bronchiectasis and acute and chronic sinusitis increased. The study also noted that IgA and IgM levels correlated significantly with these endpoints, and there was a wide variation in IgG levels in patients developing infections. Commercially available intravenous Ig formulations consist of only monomeric IgG purified from pooled plasma from healthy donors and do not contain IgA or IgM.⁵⁰ The incidence and prevalence of autoimmunity in these patients were not reported in this study.

One observational study reported a reduction in days of illness and a reduction in absence from school in 18/21 patients with PID.⁴⁵ Another observational study reported a reduction in the incidence of pneumonia (from 42/50 patients to 11/50) in patients with CVID who took monthly intravenous Ig for 6 years.⁴⁶

One retrospective study by Scheuerlein et al in Germany,⁵¹ studied 61 adult patients with CVID and auto-ITP, on IRT—of which 30 received intravenous Ig and 31 received subcutaneous immunoglobulins monthly— showed that there was no reduction in the incidence of ITP with either therapeutic strategy. Little data exist on the efficacy of IRT for autoimmune manifestations of CVID.

Although we prescribed intravenous Ig for our patient, she developed two more episodes of AIHA after initiation. The second episode was a serious event requiring hospitalisation, with a Hb of 32 g/L and total bilirubin of 40.2 mg/dL. However, she did not develop any serious infections, which required hospitalisation.

Precision diagnosis in India

Precision diagnosis is an art and science, which requires thorough history-taking, a targeted clinical examination, generating a differential list based on pretest probabilities and administering appropriate investigations with an insight into the test characteristics of every investigation administered.⁵²

This case is demonstrative of imprecision in diagnosis. For years, the patient and her family had to face the difficulties of countless hospital visits, countless poorly ordered tests and countless prescriptions of medicines, at the expense of physical, mental and significant financial strain. Several factors for these exist—lack of data continuity and provision of fragmented healthcare with each visit; lack of a primary clinician team and no follow-up; no centralised and well-maintained health record with easy accessibility. Due to such fragmented care, doctors do not know their patients well and as such their quality of work may suffer. Health literacy of the patient, cost of private healthcare and requirements of travel are some patient-related factors.

An ethnographic study conducted by McDowell and Pai⁵³ demonstrated a sustained prevalence of imprecision in diagnosis of TB in the private sector in India. The points highlighted for this imprecision can extend across all spectrums of our healthcare system:

• Poor data documentation at presentation.

• No follow-ups and poor documentation of clinical status if followed-up (7/105 doctors followed up their patients. Only 1 recorded the clinical status of his patient at follow-up).

• Need for immediate symptomatic relief and not ordering the right tests (largely due to improper counselling and poor understanding of test characteristics).

• Inexpensive and easy availability of antibiotics.

• Lack of awareness on the importance of patient education^{53 54}

Also, subsequent to poor documentation, the patient is administered diagnostic tests with each visit⁵³ and this precipitates overinvestigating, overdiagnosis and overtreatment.

In our patient too, imprecision is evident. Several episodes of infectious diseases early in her life were presumed and treated with antibiotics. Since the patient developed AIHA, a battery of poorly targeted tests was ordered, such as upper GI endoscopy, osmotic fragility test and whole-genome sequencing, which may not have contributed to the final diagnosis.

Patient’s perspective.

‘After several hospital visits and multiple admissions since childhood causing significant personal distress, life has become better with getting treated after a conclusive diagnosis was made. The number of hospital visits and admissions has significantly come down. I can now spend much more time with my friends and attend college regularly. All said and done, I’m not sure why my diagnosis was so elusive all these years and I often think of how badly we were setback financially and how much we could have saved, had a diagnosis been made earlier.’

• As communicated to us by the patient

‘After some years of physical and mental turmoil, I’m happy a conclusive diagnosis was made, and treatment administered. I’m personally relieved and happy that her medical care is now better optimised, and the frequency of hospital visits and admissions too have decreased.’

• As communicated to us by the patient’s father

Learning points.

• Autoimmune manifestations are common in patients with primary immunodeficiency disorders (PIDs). A high index of suspicion is required to make a correct diagnosis.

• Precision diagnosis and getting an in-depth insight into a patient’s history requires thorough history taking. This can be facilitated by e-logging these data as demonstrated in our case above.

• Although parenteral immunoglobulin therapy is recommended for patients with common variable immunodeficiency, not a single placebo-controlled randomised trial has yet been done to study its efficacy.

• PIDs are common are significantly underestimated and underdiagnosed as shown by data from the WHO. A history of recurrent childhood infections, atopy and autoimmunity should elicit suspicion for a diagnosis of a PID.

Ethics statements

Patient consent for publication

Consent obtained from parent(s)/guardian(s).